How American corruption turned a nuclear physicist into a redneck coder

This is a story from the “how to get into IT” series, written by an old fart, a veteran of the Brownian movement who remembers dinosaurs. Therefore, his experience of entering IT will not be useful to anyone, but is of interest from a historical point of view.

I will also share my thoughts on the interface of engineering software. While participating in the development of various software designed to speed up the development of complex systems, we periodically have to listen to complaints from new users about the “crooked and outdated” software interface. However, engineers, immersed in the problems of designing real hardware, do not ask us such questions at all, either because they have already twisted their hands about the curved interface, or it does not matter to them at all. Moreover, there are two examples when real highly professional engineers in their field made claims of the opposite nature, and the first version, the crooked version of the GUI, was more convenient, but the improvements were made by some half-pokers.

I was prompted to write this text by a conversation with one of the cool developers from a “fat” company, with whom we crossed paths on a yacht in the Mediterranean. Having learned that I was also from Baumanka and had my own business, he became interested and asked questions. How did I manage to start a business on software, why didn't I go to a big company, like Yandex, Sber and others. His acquaintance with software also began as the creation of his own development for analyzing the results of metallurgical tests in the laboratory, but ended with working as a programmer for hire. Sipping wine on a yacht somewhere between Turkey and Greece in 2023, he suggested that perhaps if he had continued writing software for metallurgical research, he could probably now be sailing on his own yacht, not a rented one, and not near Turkey, but in the Caribbean (but this is not certain). And since minced meat cannot be turned back, I decided to describe my success story, as it is funny and instructive.

Part one: beginning a career as a scientist

My first computer was an ISKRA 86 from the Smolensk plant, bought for the price of a Lada car in 1991. In the USSR, from the age of 15, I was a black market speculator, then an individual entrepreneur, selling everything including vodka and cigarettes, but the biggest business was the sale of books. But even then I suspected that computers were the future, and when my business partner bought himself a Lada from Germany, I bought myself a miracle of technology with an Intel 8086 processor and 256 kB of memory, which I later expanded to 512 kB for some money. No hard drive, DOS from floppy disks.

Its main use was to play the yoke of Pila and the “Prince of Persia”.

In 1994 I entered MSTU. Bauman to the department of nuclear reactors and power plants. And I was surprised to discover that this miracle can be programmed. The wonderful world of Pascal opened up for me, and has not closed since then. Then the process looked like this: from the beginning, DOS was loaded from one floppy disk, then Turbo Pascal and it was possible to write programs. In addition to bubble sort and spline interpolation, which were required by the Baumanke program, I wrote a program that could play the fool, using pseudographic symbols as card images, but at the same time had two levels of complexity: on the first it played fairly, on the second it took away when dealt get yourself the best cards. And in addition, I used the computer speaker for sound.

The game looked exactly like the picture, only my monitor was black and white.

The game was a resounding success with a single user who bored the entire dorm room with the obnoxious beeping of the speaker every time the cards were dealt.

I felt like a cool programmer. Taking into account the fact that there was no Internet and Google at that time, only books and the method of scientific poking were used to study programming. I was already thinking about changing my profession, leaving nuclear reactor plants and becoming a programmer.

But, fortunately for me, Windows and Delphi appeared, and that was a breakthrough. I realized that a programmer is a dead profession, if anyone can throw buttons and other components on a form, and then get a working application at the click of a button, then why study it at all? And I continued studying nuclear reactors, it was more interesting there.

In 1999 I ended up at the Kurchatov Institute. And I got there through an acquaintance, not to earn money or, God forbid, to work. By that time I already had a prestigious position as a janitor at Mosenergo. And solely in order to undergo mandatory practice. Other options implied the need to actually do something or even go to who knows where.

To avoid getting in the way, I was given a task. The boss solemnly gave me a reference book, a Soviet manual for calculating the critical outflow at various pressures from the circuits of the RBMK reactor, printed on yellow paper in an edition of “100 copies.” and asked to calculate the outflow at different pressures and temperatures for the RGK (dispensing group manifold). With the obvious hope that I will do this for the entire three weeks of practice, and he will not see me again until the conclusion of the practice is signed. But he was unlucky, I was familiar with Mathcad, already in some of my coursework, I filled it with formulas for the properties of water and even calculated a heat exchanger with a Field tube using the iteration method.

As a result, over the weekend I transferred the entire manual to Mathcad and could calculate the critical outflow from a pipe of any diameter. But my boss’s bad luck didn’t end there. The fact is that I myself am from the glorious city of Desnogorsk, where there is a glorious RBMK-type reactor that exploded in Chernobyl. And one of my first practices, also through an acquaintance, took place at an educational training center, where, as part of my work, I wrote a training manual “The RBMK reactor as a “teapot” for dummies.” To write this masterpiece, I had access to an archive with documentation on the nuclear power plant. (Those were the days: a student could go and get acquainted with any drawings of a nuclear power plant, and then put it all on the Internet. And it still does, which is typical. https://reactors.narod.ru/rbmk/index.htm). And I had a complete understanding of what pipelines the RBMK circuit consists of. Then it was a matter of technology: for each pipeline diameter found in the RBMK, a table of pressures and temperatures (also known parameters) that could exist in reality was compiled. Excel is connected, loaded into Matcad and the results are written into a set of Excel tables.

Therefore, when I showed up on Monday with the results of the calculations, the boss, of course, was very surprised, but was not taken aback and asked me to calculate the critical outflow from an individual water pipeline. There were about 10 more options in the manual, where ruptures and critical expiration were possible, and in principle, 2-3 days per option could well have taken up the entire practice. But then I told him that I had already calculated everything and turned the manual into a set of tables. The boss freaked out and took me to real scientists, of which there were two in the department. One is a master of sports in mountaineering, the other in kettlebell lifting.

It was a fun time, right after the crisis of 1998. The time when the modern Russian economy was born. As the current Minister of Defense classified Russian business: “There are three types of business: “ratting”, “grunting” and “hoofing”… “Ratting” means they swooped in, ate, and ran to another pile. “Grunting” means when there is a trough, everyone falls to it and drives each other away, and “hoofing” means, like a deer, pawing its hoof long and hard in search of a piece of lichen.”

Science at that time was also clearly divided. Those who had the right to sign simply “grunted”, those who could sell the data abroad “ratted”. I was unlucky and ended up with deer.

In Kurchatnik, the official salary was then $25. And two scientist fighters were responsible for scientific support of RBMK reactors in the country and abroad. At that time, RBMK-1500 was still in operation in Lithuania.

– Guys, I found a smart student for you, he did the calculations for me over the weekend. Give him an assignment, let him work with you.

The scientists looked at me gloomily and with great doubt, but still assigned me the first important task. They gave me money and sent me to buy vodka and snacks. I went to the nearest supermarket, bought vodka, sausage, bread, pickled cucumbers, pickled mushrooms, and a carton of orange juice. At that time, it was possible to freely carry it in bags through the entrance of Kurchatnik, through security. When I began to lay out my purchases on the table, the faces of the scientists clearly showed me that I had failed the task.

– Why did you buy so much food? We sent you for vodka, not food!

“No,” the scientists say to the boss, “we don’t need such students, let him go to Kramerov, let him calculate anything for him.”

And it must be said that Alexander Yakovlevich is a legend, the author of the brilliant book A. Ya. Kramerov, Y. V. Shevelev “Engineering calculations of nuclear reactors.” This was the pinnacle of my career as a nuclear physicist. Alexander Yakovlevich, as a scientific supervisor, instructed me to calculate the combustion of graphite in the masonry of the RBMK reactor, and he did all the main calculations and studies of sources; all I had to do was put it all into Matchcad and publish an article for which I received the Young Scientist Award. And perhaps I would have become a scientist, but then the Americans appeared with suitcases of cash dollars.

Part two, tragic: money instead of science.

At that time, the US government, in addition to supplying us with Bush legs, helped the Russian Federation with assessing the safety of nuclear power plants. She allocated a lot of money. The scheme of American assistance was approximately the same as always.

The US government allocates money to carry out calculations for an in-depth safety assessment of nuclear power plants in Russia, money is allocated to the national laboratory, at standard prices:

The competition to evaluate the reactor of the Leningrad Nuclear Power Plant (RBMK reactor), which was then undergoing modernization, was won by the cool laboratory of nuclear physics Pacific Northwest Nuclear National Laboratory PNNL. It is clear that you need to count using the American program RELAP5 MOD3 from another cool laboratory Idaho National Laboratory. But for calculations, initial data is needed. In principle, they can be taken to the Leningrad NPP, but people there have no time for that, they are modernizing, they have iron, pipes, pumps and construction. It is impossible to understand what these American scientists want. Therefore, the Kurchatov Institute was involved in the work so that American scientists would not get in the way at the construction site. Let the scientists agree in their bird language, LNPP still coordinates all decisions on modernization with the scientific director of the Kurchatov Institute, and all the data is there.

And then the American effective managers got their card, outsource. It is clear that the cost of an hour for an American nuclear physicist in 1999 is nuclear and is calculated in dollars, and compared to the salary of Russian scientists in the same 1999, it is also cosmic. Salaries there in the laboratory were approximately $3,000 – $4,000 per month, and it was at these rates that the US Treasury Department allocated money for safety assessment work. And here at the Kurchatov Institute there are professional Soviet physicists and scientists whose official salary is $25. This means that the potential margin is such that even drug dealers cry with envy and nervously smoke, or maybe snort and inject drugs on the sidelines. Americans are real capitalists; there is no crime that capital will not commit for the sake of 100,000% profit.

And here is government money for the safety of Russian nuclear power plants; only a sucker, which the Americans have never been, could not earn money here. Since I was at the end of the food chain at the time, I won’t say how much money was made from this project.

I was told a story that when an American manager was too lazy to count and give money to seconded Russians himself, he gave them a card and PIN code and sent them to an ATM so that they could withdraw money from his card for themselves.

Well, the known facts of the arrest of our Minister of Atomic Energy in Switzerland https://rg.ru/2005/05/05/adamov-arest.html there are amounts there.

One version of this story says that the underlying idea behind American funding of an in-depth safety assessment of Soviet nuclear power plants was that American physicists, using American calculation programs and data on Soviet nuclear power plants, would carry out this in-depth assessment and prove to the world that these RBMK reactors of Chernobyl like they are mortally dangerous and they urgently need to be locked up with a mother known to everyone. In addition, the RBMK of them has already exploded. An excellent plan, reliable as a Swiss watch, it would seem that what could go wrong?

But money conquers evil. And effective managers, having received money from American taxpayers, made a congenial decision – to give all the work to the Russians and pay in cash in dollars, which they leave as a tip in the canteen of the PNNL laboratory. Outsource is the basis of effective management.

As a result, the department, where there were two scientific deer, and one stupid student, unable to even buy vodka, was tasked with carrying out calculations on the American program of the Russian RBMK-1000 reactor. I had no intention of working at all, but they offered me conditions that were impossible to refuse. The boss said $150 every month, in cash. And a bonus based on the results of the project in a year. $150 a month for a Baumanka graduate in 2000, it was like now in Sberbank as a middle student. And I agreed.

The computers, of course, were provided by the Americans. They had OS/2, there was such an operating system. Now I understand that these computers most likely should have been sent to the trash heap in 1999, but they were sent to the Kurchatov Institute. The RELAP/MOD 3.3 program was also transferred. It is an excellent calculation program and has great weight and authority in the atomic field; it is practically the standard calculation program.

But if Steve Jobs had seen its interface, he would have died from an attack of botulism, plague and ectopic pregnancy. It’s good that young Jobs went to look at the software in the Xerox laboratory. If he had gone to the Idaho National Laboratory physics laboratory, there would have been no Macintoshes or iPhones.

The program interface can still be evaluated (by the way, if you go to the website of this laboratory, news from 2023):

On the speaker’s screen in the 2023 news is this picture:

You might ask, if this is the interface from the news in 2023, what did it look like 24 years earlier, in 1999? Are they really punch cards?

Punch cards? Yes, you will laugh, but it was punch cards in 2000. See the following picture – this is the actual input of initial data in the form of punched cards, which I personally used. Comments that are unimportant are marked with stars, and the entire basic description of the reactor model is a sequence of numbers and they are called cards. Yes, in 2000, when we were creating a model in American software, we discussed maps, and in the user manual these were also maps.

This is how I imagine the picture. American physics programmers are sitting there developing a program: they have lambdas, correlations, closing relations, Reynolds and Prandtl numbers and the Bernoulli, Navier-Stokes equations. They cut holes in cards with knives. An intern like me comes running and says:

– Comrade scientists, stop the stabbing! All! Progress is in the yard. Glasnost, Perestroika, Reagonomics, no more punch cards! Now you can write to tape in files without restrictions, you don’t have to worry about meaningless numbers, you can give variables names in human English. Hurray, comrades! The interface revolution that the Bolsheviks and Turing have been talking about for so long has come true!

“So,” American nuclear physicists answer the intern, “here’s a procedure for reading source data in the form of punched cards, don’t change anything there!” If the last digits in the card number are 101, then there should be diameter, length and volume, and if the last digits on the card are 103, then there should be roughness and hydraulic diameter.

– Should we set the parameters just like that?

– Yes. Like on punch cards: first the card number, then the numbers. We have already written down what the numbers mean and their order. Be careful not to get confused!

– Can variables be in text form?

– No, only numbers like on punch cards!

– Well, it’s not clear at all!

– Well, okay, don't cry. Where did they get such stupid people from?!? You can indicate the type of card in text for such morons, but no more letters, only numbers. We have fluid and gas mechanics here, Navier-Stokes and Reynolds don’t fit together and the closing relationships don’t close, get out of here and don’t stop me from thinking!

And until 1999, this is exactly the type of data entry that worked. It probably also works now, judging by the screenshot from 2023.

Armed with advanced American technology and the world's best program for calculating accidents at reactors, RELAP5 MOD 3.2, we began working for the American laboratory.

The calculation process looked like this. One of the two scientists assembled a mathematical model in the form of a set of punched cards in a text file. And he launched it for calculation, as a rule, overnight, sometimes for a day. The next day, the hard drive with the calculation results was pulled out of the sled (that was the time), and the second scientist went to the adjacent building, where there was a Windows computer with a program for viewing the results. Processing the results calculated at night took two hours. The results in the form of printed graphs were placed on the boss’s desk, in the form of graphs approximately as in the figure above. He looked attentively and asked to bring out more graphs (again, a trip to another building for about three hours). He gave instructions on what to change in the model (again on the night of calculations). If the results and set of graphs were satisfactory, the boss wrote a description of the regime and why these graphs proved the safety of the Soviet Chernobyl-type reactor. For each graph, a specially trained woman created a description and generated a report.

Since the work had already begun and was paid in cash, two scientists, despite my failure with vodka, agreed to use me. I think so, they thought that they couldn’t send me for vodka, but for running around with disks with calculations on the safety of nuclear power plants and printed graphs, even such a stupid student as I could do.

– Strong programmer?

-Then take the disc and blow into another body. Bring back a printout of the results.

But since I am very lazy, I decided to figure out what kind of file the program spits out after a day of calculation, which still needs to be gutted with special viewing tools for an hour in order to pull out the desired graph.

It turned out that the format is quite readable, although the program stupidly writes at each step of the calculation everything that it calculates for each component, and therefore the file contains a lot of unnecessary data. As a result, instead of running between cases with hard drives and printouts, I quickly threw together a Delphi program that takes the RELAP/MOD 3.2 result file and builds the necessary graphs in a couple of minutes. At the same time, you can configure which graphs and how to display them.

As a result, the boss no longer waited for a printout, but could see directly in my program what his scientists had counted on the American program.

The next step was to upload everything to text files, which I opened in Excel, where the graphs were built, and then I collected it all into a separate Word document, in which all the graphs selected by the boss were neatly listed with the content, two graphs per page. One mouse button and 15 minutes. All that was left was to add a text description and the report was ready!

It was translated into English, the details of the American office PNNL were inserted into the headers of the report, that’s it! The ass-hours of American scientists, paid for by the US Treasury Department, were no longer needed, and the money was spinning and muddying.

When our American gentlemen in pith helmets from PNNL came to us, I showed my development. It seemed to me that now the good gentleman would see my genius, how I had sped up the process of issuing reports, would pee on my shoes with delight and take me to distant America, where he would give me “a barrel of jam and a basket of cookies” and I would be a bourgeois bad guy.

But, to my surprise, my brilliant solution did not arouse any keen interest among the white gentlemen. They smiled politely and nodded, but without much enthusiasm.

Just out of curiosity, I decided to see what these guys have with their interfaces now. All the same. Photo fact: in 2019, nuclear scientists in the Russian Federation were still using homemade tools to read the results of the latest version of this cool program RELAP5 – 3D!!!! In other words, in the USA everything is more stable than ever; they haven’t been able to create a normal interface for viewing calculation results for 20 years. Ours continue to look at the results of calculations of Soviet reactors by the American RELAP in homemade programs.

Why is the interface so bad for nuclear physicists? I have two versions:

1) Professional: if the system that the engineering program considers is quite complex, then any interface that helps prepare calculated data, as well as facilitates data processing, is completely unimportant. In fact, if data preparation takes 1% of the total labor intensity of the work, then even reducing it to 0 does not affect usability in any way, and spending expensive time learning a new interface is like beer without vodka, money down the drain!

2) Economic: based on experience in implementing engineering software. From my experience, I understand why the Americans themselves will not create such a program. Everything is extremely simple: scientists are paid money for ass-hours, and the longer the program builds graphs, the more ass-hours can be attributed to the release of a report for the money of the US Treasury Department, and GDP grows. Bingo! Therefore, only stupid Russian students make programs to speed up the viewing of results and the release of reports from their American program. So, brother, you won’t sell the elephant!

Which of these two versions do you prefer?

Just now I thought, what if Matlab Simulink slows down for exactly the same reason. This is a tool for American scientists, and Matlab developers know what scientists need. When your native Simulink model takes 7 minutes to open, and SimInTech opens the same model instantly. This is not an advantage, but a disadvantage – you can’t make ass hours with SimInTech.

I got distracted again.

An auditor or “Chief, everything is lost” is coming to see us!

And just when we had set up the process and were already imagining the thick stacks of green bucks that we would soon receive, the server fox crept up unnoticed, and trouble came from where we least expected it.

It turned out that American scientists came for a reason, but to check. During the visit to our American auditors, it turned out that despite the fact that we have established the production of reports for the American laboratory on an industrial scale, we seem to have big problems with reporting. The Americans turned out to be real LGBT extremists in the bad sense of the word.

American friends asked a question as sneaky and insidious as American imperialism itself: where is your description of the reactor model?

The management provided a picture in Worde like this:

But this description of the model did not suit the American gentlemen at all. In fact, the picture had a little less to do with the description than nothing at all.

The real description of the model consisted of 5 tattered pieces of paper from the accounting drafts (when, to save paper, they are printed on the other side), on which all the elements of the model were drawn and signed in small and illegible handwriting by the author’s hands, and numbered. Moreover, a couple of these sheets were torn unevenly; apparently, they began to create a description of the model back in 1998, when there was a shortage of not only printer paper, but also toilet paper.

The main developer of the model, showing these pieces of paper, liked to repeat that there were 3 years of work here. If you lose them, the entire model will turn into a pumpkin and you can throw it away, because without these explanations the sets of numbers in the model will turn into meaningless nonsense even for the author.

It was a little embarrassing to show such leaflets to the Americans.

But the white gentlemen who arrived brought us real technology for creating such reports. It turns out that for each model that we create in the punch card style, a so-called WorkBook workbook must be created. This is a separate document that describes how, magically, from the thousands of pipelines that the RBMK reactor represents, a set of numbers in maps is obtained. For example:

For one resistance figure in the map 5.14, the American scientists had a page of two-story formulas in the documentation. And there must be two more people who created this calculation and who checked it. And this is for one parameter of one component, and there are approximately 40,000 of them in the set. Copies of drawings and equipment specifications from which the numbers were taken were filed in the same workbook. Please note that the pages of diagrams where this figure can be viewed appear to be indicated in parentheses.

The number of ass-hours paid for by American taxpayers has skyrocketed before our eyes.

At this point, two scientists who had just managed to rejoice and relax that there was no need to run between buildings, not to chase a stupid student, and the reports flew out by themselves, like from a Kalashnikov assault rifle, had a birthmark. It was clear that it was physically impossible to make such documentation with the help of two scientists and one stupid student in the foreseeable future.

Because our approach to creating the model was completely different. The model developer, a Soviet scientist, master of sports in mountaineering, took a piece of paper from the drafts, took a drawing, a calculator, a reference book for calculating resistance hydraulics from the library, lit a cigarette and drew a diagram on the draft from the printer, writing out sequential volume numbers using his own coding system. The calculated diameters, flow sections and other calculations were immediately entered into the model file with punch card numbers. No intermediate calculations, much less formulas, were written down anywhere. What about nahua? The second one was doing the same thing at that time with another part of the reactor systems.

Now you need to repeat all these steps for each number and write them down in your workbook.

The project for in-depth safety of nuclear power plants, and this is money already spent under the Gore-Chernomyrdin program, for which they are already preparing to put the Minister of the Nuclear Industry in prison in Switzerland, hangs by a thread. When asked by their superiors to make the same workbook for the American gentleman, the scientists resolutely said:

– Let them return Alaska first, then we’ll see.

– There’s a student, he has nothing to do anyway, he doesn’t even run between buildings, so let him get drunk with this crap.

I thought a little and came up with a solution. Look, if I made tables of critical flows for all pipelines of a nuclear power plant, maybe we can do the same with geometry based on the model and likeness. Let's use Idelchik's reference book instead of the method. I’ll put the formulas into Matcad, we’ll just make tables of bend diameters and pipeline lengths, and we’ll have almost a real Workbook. For each number, we will print out the formula from Matcad and the table of resistances calculated in Matcad, and select the desired value in the table. As a result, 2 pages per number, and there will be a thick document.

I ordered the Idelchik reference book from the American Amazon, and I sat down to work.

As a result, a document was born that contained formulas from Idelchik, tables of geometry parameters and pipelines typed in Mathcad+. And to create a workbook, the Matcad formula and a table were printed, in which the calculation result used in the model was highlighted. Thus, both the wolves are safe (the Americans received thick workbooks with a description of the model to justify the hours spent in front of the American Treasury), and the sheep are fed: Russian scientists who did all the work for the American laboratory received the promised dollars.

After such a resounding success, I was invited to a neighboring department, since there were still a lot of scientists at that time, and there were still few people capable of making IT that would make life easier for scientists. As a result, instead of doing nuclear physics, I began to create an interface for computational codes.

In principle, by and large, I am still doing this. And American physicists, who did not appreciate my genius, are still sitting without a good interface (but they have dollars, probably, they are fine anyway). Sometimes they lose nuclear power plants because of errors in data entry, but this is a trifle: An error in the code costs a nuclear power plant. A short video investigation of how this error appeared and how Russian rednecks offered to save the nuclear power plant:

And now, when they tell me that the interface of our program is inconvenient, I am as calm as a boa constrictor, because an inconvenient interface, after the punch cards of American nuclear physicists, does not exist for me at all.