Victory over the "dinosaurs", barter and new challenges

dev.ua continues publishing the history of the creation of IT-Enterprise, the most famous and widespread Ukrainian ERP system. Volodymyr Mikhailov, one of the founding fathers of the company, tells about the events. The next series is about the first half of the 90s.

How "tickets" defeated mainframes

And here we pulled out another lucky ticket, which was the attitude of 90% of IT departments of large enterprises to this technology. Every industrial enterprise already had information and computing centers with large IT departments that had been writing programs in PL/1, COBOL, Fortran for years... A lot was written, all this had to be supported, and there was neither time nor desire to switch to new technologies. Therefore, they took personal computers "at bayonets": they said that these were frivolous trifles, and real programmers only work on mainframes.

For us it was Bingo!!! Jackpot!!! Because we knew how to program on "things" and knew the needs of industrial enterprises. We could go in and sign development contracts with virtually no competition.

Reflecting on these years, I see that it was the development of individual tasks for the client. At the end of the project, it was something like this: “We did it, thank you, goodbye.” Why is that? We worked with industry, which already had large automated control systems, and we were ordered only those tasks that they lacked, mainly in production management. The mainframes were still working, the ACS departments were also working and were holding on to the “warm pipe of their native plant” with all their hands. The work with the Donetsk NORD Refrigerator Plant was very indicative. After the survey and the proposal to develop a production management system and feasibility planning for them on a network of personal computers, we were pushed out of there by our colleagues from the IT department, after the head of the department
The ACS realized that all of its “50-year-old girls” computer operators would simply have to be fired because they were unnecessary, “and he had lived his life with them.” That's why we grabbed onto developing anything.

After 5-7 years, we gradually took away all management tasks from several customers, implementing them on a network of personal computers. "Suddenly" it became clear: mainframes were no longer needed. Everything that worked on them was already being launched in parallel by us on new platforms. The end for the "dinosaurs" came inevitably - the disbandment of old IT departments, layoffs, recruitment of new specialists. Ignoring technical progress is like death! But that was later...

The Horrible 90s: Meat, Refrigerators, and Underwear Instead of Money

When 1992 and 1993 came, the planned economy collapsed, and there was no other. Industry was at a standstill, there were massive non-payment of salaries, unemployment, and rampant inflation, which was fueled by the populist decisions of the authorities. In fact, during these two years, the
Most of Kyiv's scientific institutes, in particular the Institute of Cybernetics in Teremki, were virtually depopulated because there were no state orders "for science," and the Kyiv military-industrial complex simply ceased to exist.

By the way, it was this year that I first noticed the young Viktor Yushchenko, who, standing on the rostrum of the Verkhovna Rada, simply plowed into the deputies, trying to convey to them that the law on indexation of all salaries and social benefits, which they approved for the Nth time in a row, would only lead to even greater impoverishment of the population. But the law was adopted for the Nth time and inflation reached 1000% per year. Under such conditions, development contracts concluded at the beginning of the year were already worth several kilograms of meat in April. That is, money ceased to be money and “barter” reigned in the economy. We submitted the project, and in response we were told: “There is no money. We can pay with products.”

Horror! And what to do with their products? We are programmers, not traders. For example, the Donetsk Refrigerator Plant paid us with its Donbas refrigerators. One of them worked at my house for about 15 years. The Radomyshlanka Garment Factory paid for several hundred tons of bed linen for a production accounting project. There was so much of it that our families used it for the next 20 years. Back then, no one could even imagine what awaited us and what fate this linen would have.

In August 2014, when Ilovaisk happened, my wife and I joined the volunteer movement, came to the Main Military Clinical Hospital on Lesya Ukrainka in Kyiv and asked what we needed. The head doctor of the surgical department wrote a list of all kinds of medicines, which we bought in pharmacies and brought to the hospital. Suddenly, the doctor said that if you have old bed linen, you can bring it - there is not enough. And then we realized that 20 years after the calculation for the automation project of the early 90s, we still had 25 kilograms of brand new sheets, duvet covers and pillowcases left at home. We brought all this to the hospital - the doctors were shocked and burst into tears from such wealth.

They paid us everything in the early 90s. Cable, chandeliers, meat... It's funny now, but those were the times, and there was such "money."

How we covered all production tasks and got new roles

The drop in PC prices allowed us to buy a second computer and take on more work. During those years, we were able to handle all of our clients' requests. With a RAD development system, we were able to do it all on time and make good money. Looking back, I see that projects like these allowed us to build
a full-fledged technology for quickly creating software products and practically step by step covering all possible management tasks at enterprises.

I spent several years in awe of the ability to develop anything and launch it on a client's computer very quickly, in the imagination of those years. I felt real satisfaction from watching how clients use your software.

Did any doubts arise in my head, whether I was doing all this correctly and doing the work that I had dreamed of since childhood? Considering the fact that we secured for ourselves an income 20 times greater than that of the researchers around us who were sitting next to us on the fourth floor of building 18 of the KPI, no. The thought in my head was “I dreamed of being a programmer, so I became a programmer, that is, I develop programs and satisfy the client — I write for him what this client needs. And this brings in a lot of income. Everything is true.”

And that's exactly the idea that prevailed in the world - development to the order of each client. That's how SAP began - its founding fathers worked in the European branch of IBM and created custom software for each customer.

How programmers outgrew pure programming

Gradually, around the age of 27-28, I started to get bored with writing programs. By this time, I had mastered 11 programming languages, from two IBM and DEC assemblers, through all the standard mainframe languages ​​to GPSS, StamClass, Pascal, C, Natural, Clipper, Foxpro, …. Doing the same thing ten times in a row, but with different syntax? Then it wasn’t so interesting anymore…

I always asked myself the question, what will happen in 5-10 years? What will happen in a year no longer worries me, it is clear that it will be the same as now, but, probably, a little more...

Within a few years, clients and even colleagues began recommending us to their friends, and we had a steady stream of orders. We needed to expand and hire specialists.

Unfortunately, and now I understand that fortunately, the absolute majority of young specialists and senior students of computer specialties of the Faculty of Information Technology of the Kyiv Polytechnic Institute, whom we invited to work, were fantastically gifted programmers, but with minimal skills in communicating with customers. If for me, even at the age of 23, there were never any problems in communicating with customers (to come to a huge enterprise and examine the subject area while communicating with many heads of various services, to agree and defend my decisions, to sign technical documents, to hand over the project to the customer, having withstood the "usual face-off during acceptance", to conduct user training...), then the absolute majority of young programmers trembled, sweated, fainted and... in the end, they were relieved from the very thought of the prospect of communicating with
"fat uncles-customers." Therefore, when recruiting new programmers, we had to take on all this work ourselves.

Over the years, we quickly realized that we shouldn't write all the programs ourselves with Oleg Shcherbatenko, but rather specialize and fulfill the roles of project managers, business analysts, and designers. Provided that we have properly scheduled programming task statements from us,
programs can be written by "pure programmers" who did not want to understand the subject area at all and were even proud of it. To do this, it is necessary to implement programming standards, typify interfaces and generally unify information processing schemes. This achieved complete substitutability
programmers, absolute standardization of interfaces across all programs, and minimized training time for both new programmers and end users.

Super-efficiency was also achieved because, when designing new tasks, I understood down to the last program command how it should be written, that is, the most qualified programmer became a business analyst and designed what needed to be programmed, understanding how to do it in detail.

At this stage, my role changed - now I am no longer a programmer, but a project manager, designer and the most qualified developer. Now it would be called a PM, architect, team leader, tech leader, business analyst and senior manager all in one, but in the early 90s everything was simpler.

RAD is a superpower, but it has become the standard for everyone. We need to be different.

The term RAD (Rapid Application Development), which appeared at that time from the light hand of James Martin in his super-popular book of those years “Principles of Software Development”, was remarkably fully implemented in those years in FoxPro 2.0, which became our main platform, and in several other platforms such as PowerBuilder, Delphi…

And if in the early 90s, programmers around us, who had the richest experience of programming on previous platforms, tried to write in the old style, and we took advantage of the fact that we could develop software at least 10 times faster with the help of RAD, then in a few years everyone around us
We mastered the new tools and our competitive advantage gradually evaporated. We had to be different from the others in some way. But more on that in the next series.

Post Scriptum from 2026

Drawing analogies with the beginning of 2026 and the euphoria from the AI ​​frameworks for creating software that have appeared in recent months and the slogan "programmers are not needed," I predict that this new spiral of development will be mastered by the entire IT community in a couple of years with global changes in the structures of IT companies.

At the same time, those IT companies that miss this new branch of development will disappear. Most likely, they will not disappear yet, but will remain only to support old solutions and still disappear a little later... And the arguments of rejection from "experienced developers" who are gurus for previous ones
technologies, will be the same as they were 30 years later: “this is not programming, but who knows what”, “I cannot be responsible for code that I did not write myself, it is some kind of nonsense”, “this code lacks some features that are in our old code, and customers need these very features-bows”, ... and many such “arguments” that the market will not accept.

In 2026, I expect disruptions and market changes.