Contents:
- Anton Zhiyanov
- Most popular libraries are open source
- About the most stellar project on GitHub
- Under what license should projects be hosted on GitHub
- How much effort and time goes into maintaining open-source projects
- What's wrong with Open Source and why corporations are flirting with it
- How to become a maintainer and where to start contributing to Open Source

Free Course: "Quick Start in Python"
Learn MoreAnton Zhiyanov

About the expert
The expert has deep knowledge and experience in his field, which allows him to provide valuable recommendations and solutions. His qualifications are based on many years of practice and continuous training, which makes him a reliable source of information. The expert actively shares his knowledge through publications, seminars and consultations, helping others achieve success. His approach is focused on the individual needs of clients, which ensures a high degree of satisfaction and trust. Thanks to his analytical thinking and ability to assess situations holistically, this expert is able to find optimal solutions to complex problems.
He is a technical lead at DaData on Twitter, @nalgeon. He actively uses Python and SQL, sharing his knowledge and experience in his publications. On Twitter, he discusses various aspects of programming and development, creating open-source projects and training courses. He also writes for the Habr platform and maintains his own website, where he publishes useful materials. Among his open-source libraries, iuliia-py and sqlean stand out, having gained recognition in the developer community.
Links play a key role in the structure of websites and their SEO optimization. They provide navigation between pages, helping users find the information they need. Furthermore, high-quality external and internal links help increase a site's authority in the eyes of search engines. It is important to use anchor text correctly so that it matches the content of the landing pages. It's also important to monitor the quality of your links, avoiding links to unreliable sources, as this can negatively impact your ranking. Link optimization is a crucial aspect of any strategy to improve your website's visibility in search engines.
I work at DaData, a cloud service that enables businesses to obtain customer information using data such as their tax identification number (TIN) or address. Our services are especially popular among online stores, allowing them to minimize the amount of data requested from customers. Customers only need to enter the delivery address, and the store receives the necessary coordinates, information about the city district, and the nearest metro station. This is especially convenient when interacting with legal entities, as instead of 20 fields with details, the client only needs to provide the company name or TIN. DaData ensures simplicity and efficiency, which contributes to an improved user experience and increased conversion for businesses.
In my professional work, I actively use open-source libraries, including my own developments. My GitHub account has been around for many years and contains several projects, including one with over 1,000 stars. In this article, I'll share my experiences working with open source projects and highlight the main challenges developers face in the open source community.
Most Popular Libraries Are Open Source
Open Source is the primary way to distribute libraries today. Commercial libraries are typically significantly fewer in number and are mostly found in highly specialized areas. Open Source offers developers access to a wide range of tools and resources, which fosters innovation and collaboration within the community. This makes Open Source the preferred choice for many projects, providing flexibility and customization.
Creating an open source library has its advantages. On the one hand, other developers can contribute, find and fix bugs, which helps improve the quality of the product. On the other hand, there's a sense that you're doing good for the community by allowing users to adapt a library to their needs rather than being limited by the author's original intent. However, I believe that not all software should be open source. Open source is an effective way to distribute libraries and some other software solutions, but it's not a one-size-fits-all solution for all software.
Last year, I created my first major project on GitHub related to transliteration. Transliteration is the process of transcribing Russian words using the Latin alphabet, similar to how names and surnames are written in passports. For example, if you need to send a letter from the United States, the address will be written in Latin letters, not Cyrillic. My GitHub project offers a convenient tool for this task, simplifying the process for users who need to work with Russian texts abroad.
Transliteration is a complex task, as there are 20 different standards, most of which do not take into account the specifics of the Cyrillic alphabet. If your first and last names are simple, you're lucky—they'll be transliterated correctly in your passport. However, if your name contains the letters "e," "yo," "yu," "ya," or the "y" sound, it may cause difficulties. Yulia is considered one of the most challenging names to transliterate.
I've developed a word transliteration library that supports 15 different schemes. The library is called iuliia, after a popular name. It's available not only for Python but also for JavaScript. I've also created a website with a detailed description of the library and demo examples. All transliteration schemes are compiled in JSON format, making it easy to adapt the library for use in other languages, such as Java.
iuliia's JavaScript version has only had 111 downloads in the last week. This is because transliteration isn't a highly sought-after programming task. Users have reported bugs several times, but no pull requests have been received. This isn't surprising, as transliteration is difficult to implement new features.
Later, I published an article on Habr, detailing the complex challenges the Cyrillic alphabet faced in the 20th century and the impact of transliteration on its development. Readers actively expressed their experiences in the comments, sharing personal difficulties related to the spelling of full names. This article attracted the attention of many developers, who then created more than 10 different transliteration solutions in various programming languages.
About the most starred project on GitHub
I have a project called sqlean, which has received 1000 stars on GitHub—more than the sum of the stars of all my other projects combined. Before delving into the details of this project, let me make a small lyrical digression.
The SQL language occupies an important place in data analysis, providing a convenient and effective tool for working with information. It surpasses Excel in functionality, while remaining more accessible compared to Python. SQL is a declarative language, which allows users to focus on the results they want to get, rather than on how exactly the database management system should execute queries. This makes SQL an ideal choice for analysts and data scientists seeking fast and accurate data retrieval.
SQL and its constructs remain relevant and time-tested. The language changes less frequently, and new constructs do not replace old ones, as happens in front-end development. Many tasks that developers solve with Pandas are performed much faster and easier in pure SQL. SQL offers powerful tools for working with data, making it indispensable in analytics and database development. Given SQL's stability and efficiency, it remains the preferred choice for a variety of data processing tasks.
Among relational database management systems (RDBMS), I particularly prefer SQLite. This compact database is easy to embed into code, making it ideal for data analysis and use as a storage medium for small applications. This is why I write articles about SQLite and actively strive to popularize its use.
SQLite has limited functions for working with text, numbers, and mathematical statistics. Although developers add new features over time, their number remains insufficient compared to PostgreSQL. This can be a limitation for users who require more complex operations and analytical functions. Therefore, when choosing between SQLite and PostgreSQL, it is worth considering the functionality and scalability needs of your project.
Third-party developers create extensions to add missing features, such as working with regular expressions. Regular expressions allow you to search and replace text not only for exact matches but also for complex patterns. These tools are available in all popular database management systems except SQLite. To integrate regular expressions into SQLite, you need to find a suitable extension, compile it, and properly connect it to the database.
For most users, the compilation process remains a complex task. Many people come to the SQLite support forum with questions about how to add a specific feature, and are prompted to download and compile a C extension. Confused, users ask, "What is compilation?" and are given a link to the documentation. Probably only a few achieve successful compilation.

SQLite extensions are often scattered across GitHub without a clear structure and logic. I organized them by problem-solving approach and created documentation so users could quickly and easily download the necessary packages for Windows, Mac, iOS, and Linux. Users sometimes suggest new features, and I had a contributor develop a module for working with IP addresses. He didn't simply copy-paste existing code; he built it from scratch, which highlights the quality and uniqueness of his work.
My work has become popular because users prefer to avoid the complexities associated with compilation. While it's possible to study the documentation and understand the process, most people don't want to delve into these technical details. In fact, it's quite simple, and I strive to make this process more accessible to everyone.
Later, I published an article about the library and posted it on Hacker News. It reached the top spot on the main page and became the most popular article about SQLite in the platform's history. Hacker News is similar to Habr, but is aimed at an English-speaking audience, providing links to articles instead of the full text. This resource is extremely popular, regularly hosts active discussions, and is read worldwide.
Which license should I use for hosting projects on GitHub?
For projects on GitHub, I prefer to use the MIT License because it is simple and clear. This license requires users to attribute the authors and not hide their names. At the same time, the MIT License is broad: the code can be modified and used commercially. This makes it an ideal choice for developers who want to make their software accessible while maintaining their copyright.
Unlike traditional licenses, open-source licenses require all derivative software to remain open source. This licensing model is impractical for libraries. For example, if you're developing a large system and use a small library licensed under the GPL, this may require you to disclose the entire source code of your development. This approach limits developer options and makes using libraries with open-source licenses less attractive for commercial projects.
I view licenses and open source from a practical perspective, so the debate between the FSF and OSI doesn't interest me. Sometimes proprietary software is necessary, too. That's why developers have adopted the concepts of Open Core and delayed publication, where software is initially released in a proprietary version and then gradually becomes open source. These approaches have their place in certain situations, and one can't categorize licenses as either good or bad. However, if you present a project using Open Core to Richard Stallman, he probably won't be able to resist a negative reaction.
How much effort and time goes into maintaining open-source projects
I invest minimal effort in maintaining my projects, as I develop them without external dependencies. The functionality of my programs depends entirely on the standard library, which is highly stable. This allows me not to worry about frequent updates and new vulnerabilities associated with third-party libraries. This approach ensures the reliability and long-term sustainability of my solutions, which is especially important in modern programming.
I don't like the JavaScript world because of the large number of external dependencies. As soon as a new project starts, it begins to become outdated within a few months. This forces developers to constantly update libraries, which often turn out to be incompatible with each other. Regular updates are becoming a necessity, which creates additional challenges in project maintenance.
In sqlean, the focus is on adding new features, and updates to older versions occur only when bugs are discovered. Problems most often arise with more complex applications or those that interact with external APIs. For example, if a library uses a cloud service, it must be updated whenever the service's API changes. This requires a careful approach to support and updating to ensure stable application operation.
I work on projects in short intervals. Regular support requires a few hours a month, while developing a new extension for SQLite takes from one to several days. I can spend weeks on new projects, which allows me to dive deeper into each task and achieve high-quality results.
Contributor activity statistics are available on GitHub, presented in the Contribution activity section. Each year is represented as a square: the more code the developer has contributed, the brighter and more saturated the square's color. In my case, due to the irregular nature of my work on projects, the colored squares alternate with gray ones. This visualization allows you to quickly assess the developer's involvement and contribution to projects, which is an important aspect when seeking collaboration and assessing professional skills.
What's Wrong with Open Source and Why Corporations Are Playing with It
Open source is an ideal solution for libraries, so most of them are openly licensed. This makes it easier for developers to detect and fix bugs in projects and also provides users with a wide choice. Thanks to open source, you can find almost any library and adapt it to specific tasks. Using open source libraries helps speed up development and improve software quality.
The code distribution model has two serious problems, both of which are related to issues of fair profit sharing. These issues significantly impact the sustainability and effectiveness of this model, requiring careful consideration when implementing it.
Many libraries are used widely, including by giants like Google and Facebook. However, their creators are not adequately compensated for their work. When your developments underpin a vast amount of enterprise software sold for millions of dollars, while you yourself live in humble circumstances and barely make ends meet, it evokes a feeling of profound injustice. It is important to recognize the value of open source and its impact on technology development, as well as the need to support developers who put their efforts into creating useful tools.
The situation in the library development industry is gradually changing. Some authors are already earning between $5,000 and $10,000, which is a decent salary for an individual developer. However, if a library is being developed by a team, working for a commercial company becomes a more profitable option. Despite this, such authors are still few and far between. Most developers receive only symbolic compensation or are left without income at all.
Much depends on your expectations. If you plan to develop open source libraries for profit, be prepared for potential disappointments. It's important to recognize that success in this field isn't always tied to financial results. Set aside your expectations and approach the process with an open mind.
Open source raises many issues related to social justice. Companies like Google may provide higher compensation to authors of popular libraries they use. This creates an inequity between developers who make significant contributions to open source projects and large corporations that profit from these technologies. It's important to discuss and find solutions for equitable compensation for developers in the open source community to ensure sustainable development and support for innovation.
Popular programs like PostgreSQL and Sentry were originally distributed under permissive open source licenses. This allows corporations to offer open source products as cloud services. As a result, Amazon significantly increases its revenue, while independent developers lose users. Many customers prefer large, trusted companies like Amazon, which can offer better terms than smaller companies.
Some companies are migrating to the Open Core model, while others are opting for new licenses, as Elastic recently did. Elasticsearch is one of the best tools for searching various types of data. As a result of Elastic's abandonment of its permissive license, cloud services were prohibited from reselling Elasticsearch. This caused a backlash, and Amazon forked the latest open-source version of the program, promising further development. Amazon has significantly more resources than independent developers, which puts Elasticsearch in a vulnerable position and could threaten their business.

The developers also encountered a similar problem. MongoDB clearly stated in its license that their software cannot be freely resold on cloud services. As a result, Amazon was forced to negotiate payments with the company to ensure the ability to provide this service to its users.
Cooperation between corporations and open source often leads to mutually beneficial results. Google invests in open source projects for pragmatic reasons: it uses these programs and libraries in its services and seeks to adapt them to its needs. Therefore, large businesses are interested in influencing the development of open source software. In turn, the authors of open-source projects receive funding from contributors. This cooperation is based on mutually beneficial interests, not altruism.
How to become a maintainer and where to start contributing to open source
According to statistics, reaching 1000 stars on GitHub is a very difficult task. To avoid disappointment with Open Source if you fail, it's important to set expectations outside the process. Success in this area takes time and effort, so it's best to focus on creating quality content and actively engaging with the community. Don't search for an idea. If you have a problem that can be solved with code, create a library or utility. After that, you can consider keeping the project for yourself or formalizing it, writing a minimal readme and publishing it on GitHub. It will take a little time, but you'll create something valuable for the community, which will increase your level of participation in Open Source. You'll not only solve your problem but also help other developers, which can lead to new opportunities and interactions. The easiest way to start contributing to projects is with the libraries you already use in your work. This will allow you to not only improve your skills but also provide real assistance to the community. You're already familiar with the functionality, which makes it easier to learn the code and understand how to make changes. Start with small fixes or adding documentation to get used to the process. Later, you'll be able to move on to more complex tasks and significantly improve your development skills.
The first method: Go to the project's GitHub repository and review the open issues. If you can solve any of the problems, submit a pull request with your fix. However, contributing isn't limited to just fixing bugs or suggesting new features. You can also create documentation for the project, as open source projects often lack it.
The second method is to identify a problem in the project yourself and file an issue. Some maintainers will appreciate your initiative, while others may not be interested, especially if they already have many open issues. To increase your chances of a positive response, it's recommended to immediately attach a pull request with your proposed fix. This will not only make the maintainers' job easier, but also increase the likelihood that your message will be heard.
Avoid submitting pull requests containing 10 files with 2,000 lines of changes. If this is your first time contacting the maintainer and presenting a large amount of code, your pull request will likely be rejected. Instead, create an issue detailing the problem and proposing your solution. You'll likely be asked to make more localized fixes rather than rewrite the entire code.
Small changes can be submitted immediately, while large changes are recommended to be discussed with the maintainer beforehand. This will save time and avoid a situation where your work is rejected, which can be frustrating. Discussing major changes will help you better understand the project's requirements and increase the chances of your pull request being accepted.
Let's summarize.
- The Open Source model is great for licensing libraries, but not all software needs to be open.
- As a rule, the MIT License or another permissive license is suitable for open-source projects. It obliges the user to indicate or not hide the name of the program's author, but otherwise provides freedom of action.
- The problem with Open Source is the unfair distribution of income from programs. Most authors of popular libraries, which power the entire Internet, receive pennies for their work.
- To become a maintainer, you don't need to specifically look for an idea for a project. Try writing a program that solves your problem - in the future, it may turn into an excellent open-source project.
- Contributors can contribute in two ways: by solving an issue from the list or by finding a new problem and proposing a solution. You can fix bugs, propose solutions, and write project documentation.
Read also:
- How the FOSS model penetrated business, science, education, and creativity
- "Normal programmers don't code in Russian": 4 annoying myths about the 1C language
- Crawl like an ant, fly like a bee: algorithms invented by nature itself

