Closed software as a way to rob the state. Government procurement and import substitution
In short: free markets and private companies can work well. And another model is when he buys a state at his own expense and also makes a state. The department where I studied was associated with NIUMS – this is a state IT department where 7 thousand people worked in Soviet times and they developed software for a bunch of enterprises. But when software is purchased at government expense, and the profit is private, the situation is even worse. A state account can be a budget (federal/municipal), a unitary enterprise, or even a joint-stock company completely or mostly controlled by the state. Which the state forces you to purchase software. But the trouble is that much of this software is prohibitively expensive. And in terms of quality it would not be entirely competitive in a free market. It even gets to the point that “In the near future, state-owned companies may be prohibited from developing software for their own needs, without further selling it, if it already has analogues on the market. The Ministry of Digital Development is preparing new restrictions so that the state segment will bring its money to software developers.” It’s an unhealthy situation when the expenses are public (state), but the profits from the sale of software are private. It is possible to understand the logic of the state, but in the conditions of a developed competitive market, the factor that would force one to do 1. Qualitatively 2. By including adequate profitability in prices disappears. State corporations don’t just cut their own software. We have on the market both solutions that were successfully sold abroad until 22 years ago, as well as “crafts” that people buy based on the fact that they are forced and when working with them, the feeling arises: “This is the developer company that should pay us for the fact that we act as testers. And we work as analysts who explain what functionality large businesses require. And we struggle with glitches, learning to work around them.”
And it would be nice to spend money at state expense only if the licenses are exclusive and the source codes are transferred to the state fund. Having developed a system for one region, another can take it and modify it (by a vendor, contractor or on their own) to suit their needs.
Details:
One of the cardinal sins when introducing IP, in my opinion, is to put the customer (for whom you work) on the needle [неконкурентную] supplier. I worked for several years in an enterprise subordinate to one of the regional mines of trances and experienced a burning sensation in the lower back due to the costs that the enterprise and the budget incurred. I really wanted to cut them. And he would increase the anemic efficiency of the system for which I worked. However, they explained to me that you are counting money, this is not the budget of the enterprise and the region. Then they explained that, due to the peculiarities of the system of spending budget money, high authorities are not especially praised for the savings, but the budget for the next year will be cut according to the logic: the money planned has not been spent, which means that so much is not needed. Let's remove it. I ended up quitting. And now in another region in one of the large PJSC.
I implemented and supported, and sometimes even modified, one information system for accounting for fares. All residents of large cities are familiar with such systems. The conductor in a terminal with a terminal similar to a bank terminal (or often directly with a bank terminal) accepts funds from the passenger, a ticket is issued to the passenger, and the payment information is processed through a specialized application server. Usually, variability is allowed – the terminal may be separate from the conductor – it weighs on the handrail, perhaps next to the turnstile, and controllers check the payment (before this, the subject of the federation must set a fairly high fine for ticketless travel – otherwise there will be many “hares”). Payment is usually possible in several ways – cash, bank card, preferential travel card (students, pensioners, disabled people, veterans, Chernobyl victims), general civil travel card (with or without a limited number of trips. After processing, reporting is generated on the basis of which the work is accounted for for the transportation of passengers and often the distribution of funds. For example, according to the transferred benefit, the state pays the lost profit to the carriers. If the carrier transports citizens for money, he can also receive a subsidy – let’s say there are 10 people living in the village, from a business perspective, it is unprofitable to drive a bus there every day. But from the point of view of citizens and the state, it is also impossible not to go there – this will limit the freedom of movement of citizens (not everyone has a car or even a license. City routes are also unprofitable. The fare is the same everywhere, even when a passenger travels 500 meters in the city center). and when people are transported 30 km from a remote area, the cost of transportation differs dramatically. There are routes where they travel from one terminal station to another (minimum earnings), and there are routes where 3-4 people travel at one place during one flight (that is, profit per flight from each occupied seat increases multiple). Therefore, such systems are often purchased by the state. And the state also operates – it takes into account transport work and gives subsidies to carriers.
In this case, how can a system supplier rob the state (at the same time, there may also be beneficiaries on the state’s side – but the topic here is not entirely about corruption)
The low price of the system itself and the impossibility of its full operation without the participation of the developer company. In this case, the purchase is possible at a price that does not require procurement by auction or competition under 44 or 223 Federal Laws, but then the state will pay a lot of money for operation – and due to the presence of an already purchased system, the operation of which can only be supported by a developer with this developer and a contract for system support is concluded – and there the amount can already be very large).
Protocols for interaction between parts of the system are undocumented. And only part of the system components is sold. Let’s say that only a reporting system is purchased and the operation of other systems is purchased as a service from the “needed” company from which they were purchased. Even if all the server software is transferred, there remains quite a lot of relatively legal money from the state budget (or from other market participants). Let’s say software running on CVTs can be purchased by a region conditionally unlimited for the entire territory of a federal subject or city. Or it can be sold to carriers – the only LLC authorized to work in the region. Or maybe they’ll even squeeze it out for rent – a temporary license, let’s say a couple of thousand rubles from one CVT per month, with three thousand buses it will give a flow of 6 million rubles. Even in one year, and in the case of software development from scratch, the amount will be significantly more than the cost of development and support. Calculate for yourself how many developer man-hours can be bought for 72 million. It is impossible to develop alternative terminal software because the protocol for interaction with the server is closed. Replenishing travel cards is also a source of income. A software license for replenishment points can also be a good source of income. I’ve also come across options – someone takes a fee from the organization actually involved in replenishment (for using the software), and someone creates a desk of one and a half people, which alone has a license for the required software and which therefore wins the competition to take a percentage of each replenishment travel card, while in fact the work will be performed by its subcontractor
There are two more elements of the system that can make the owner dependent on the supplier: the format for recording data on the card (so-called mapping information, where and what is stored and how it is encoded) and the encryption security keys or the algorithm by which they are generated and the information necessary for its operation ( let's say salt). Otherwise, you may learn that when you change the system, it will be necessary to replace the cards of all passengers (and this could be hundreds of thousands and issuing cards is another possible waste).
I believe that this class of systems is not unique in its kind and when purchasing and operating other systems there are similar subtle points
How would I prefer?
I personally would include the following requirements in the requirements for state (and near state) procurement in software:
Openness and documentation of flows and interfaces in the interaction of system components
ALL custom development with transfer of software source codes. Moreover, the source texts should be accumulated in a repository that is accessible to all state enterprises. This will avoid situations where different regions buy the “development” of similar software at government expense. At the same time, the supplier company essentially sells only minor improvements to its system. If this approach is implemented, the software itself can either cost almost nothing (it’s ready-made), or it costs much less; competition will be created to improve this software; improvements will also be possible by local specialists of the customer’s company. Look at 1C. With all the shortcomings of the platform, I see a big plus – when buying solutions, I can contact the developer company for modifications (the default option), but if the price is inadequate, I can contact any of the 1C franchisees or frequent freelancers (competition), and if modifications are regular and Their volume is distributed approximately alone – you can take your own 1C nickname – it will be even more interesting based on the cost of an hour of work
When purchasing systems, analyze not only the cost of the system itself, but also the necessary expenses for its launch, commissioning and others during the life cycle. Both in terms of additional systems developed to work as part of it but not purchased, and in terms of other software – for example, you can purchase for 50 thousand a system using Oracle as a DBMS, but in the procurement documentation it will be exactly 50 thousand, although the costs state or a specific government organization will be many times higher. And if there is a system competing with it for 100 thousand that uses postgress, it is more economically feasible to buy it. Within the framework of Federal Law 44/223, all this is not taken into account. I am familiar with a situation where software was purchased for 99 thousand (to avoid competition and questions from the FAS), but in fact it was almost a cloud client; millions were paid annually to the seller of this software.
When purchasing or ordering a specific information system, it is necessary to involve specialists with experience in purchasing and using just such systems at other enterprises. There are too many non-obvious moments that cannot be predicted at first.
ps I’ll add based on the results of this and last year and current import substitution: I’d really like to see internal GNU for system-wide. Not Astra Linux and p7 office for a tidy sum, but something that will be open and accessible. If the goal is import independence, and not the earnings of a handful of LLCs selling their handicrafts to large businesses and the public sector (ordered to switch to domestic – a task from the very top), then it is reasonable to create a state institution that would adapt to domestic needs, libra office / OO. Well, or if you really want to start from scratch, let them write a new vocabulary. In the USSR, the public sector coped with such tasks. Of course, this change will require changes to the current system: a state enterprise often cannot pay a competitive salary to an employee. But it’s easy to pay 2 times more to a private contractor LLC for the same work. For additional money, technical support and adaptation to the needs of a specific customer. The main thing here is not that this is a state-owned enterprise, but that the result is not private. Otherwise it turns out sad. The state has not erased expenses from the public sector (state organizations, government entities, state unitary enterprises, various LLCs and joint-stock companies which are entirely or entirely owned by the state). The business also bears expenses (and this is money that it could spend on something else, even taking it as a profit and paying taxes). At the same time, what we are replacing imports with is not always of a high level, to put it mildly. Conventionally, Kaspersky fine reader or Acronis competed quite successfully in the foreign market with foreign analogues. Alas, Clean is of high quality and not expensive, in many areas, we are unlikely to see the price – the market is too small, and the cost of one installation is a high share of the developer’s expenses. Well, I don’t see any importer substitutes in the private sector who are willing to work at a loss (or with a low mark). It’s better to take GNU and finish it. Subsidizing at the expense of the state and, more broadly, the public of private companies does not appeal to me at all. It turns out that the costs are public. And the profits are private.
