GitHub has published a report analyzing statistics for 2023. Key trends:
- In 2023, 98 million new projects were created on GitHub (compared to 85.7 million in 2022, 61 million in 2021, and 60 million in 2020). The total number of projects reached 420 million (+27%), while the number of publicly available repositories was 284 million (+22%).
- The overall contribution of participants to all projects is estimated at 4.5 billion actions (commits, issues, pull requests, discussions, reviews, etc.). For comparison, in 2022, there were 3.5 billion such actions.
- The GitHub audience grew by 20.2 million users (+26%) over the year, reaching 114 million (last year it was 20.5 million, the year before that 73 million, and three years ago 56 million). The United States ranks first in the number of developers, followed by India, China, Brazil, the United Kingdom, Russia, and Germany.
- About 65,000 new machine learning projects were added over the year, which is 248% more than last year. The number of participants in AI projects increased by 148% over the year. The main trend of the year is the growing interest in developments related to artificial intelligence technologies. According to a survey conducted on GitHub, 92% of developers are already using or experimenting with AI tools.
- JavaScript remains the most popular language on GitHub. Python holds the second position. TypeScript claimed third place, pushing Java to fourth. C#, C++, and PHP secured fifth, sixth, and seventh places, respectively. C moved up to eighth place, displacing Shell to ninth. Go occupies the tenth position. Ruby, which previously held tenth place, has dropped to fifteenth. Significantly (+36%) increased popularity was observed for the domain-specific language HCL (Hashicorp Configuration Language), which took eleventh place. Kotlin, Dart, and SCSS took twelfth, thirteenth, and fourteenth places.
- Rust leads the ranking of the fastest-growing languages (+40%), followed by Lua, TypeScript, and HCL. An increase in popularity was noted for T-SQL and TeX.
- The number of jobs performed in GitHub Actions increased by 169% over the year.
- More than 80% of changes are made in private repositories. There were 4.2 billion changes in private repositories and 310 million in public repositories.
- The popularity of applications designed to run in cloud environments (cloud-native) or that use the 'infrastructure as code' (IaC) model based on Git is growing. More than 1 million public repositories for container creation utilize Dockerfile.
- The number of developers using automated vulnerability tracking systems in dependencies has increased. The number of accepted Pull Requests from such systems rose by 60% over the year.
- Over 301 million changes were accepted in open projects on GitHub over the year.
- Leading in number of participants are the repositories for VSCode, Home Assistant, and Flutter:
- In terms of new participants joining the development, the leading repositories are:
At the same time, the winners of the GitHub Awards 2023 have been announced:
- Nomination 'Wonderfully Welcoming': Python Software Foundation.
- Nomination 'Global Grandiose': Storybook, a web framework for building interfaces.
- Nomination 'Hardware Hacker': Limor Fried (@LadyADA), leader of Adafruit Industries, which produces open hardware.
- Nomination 'Noteworthy Newcomer': LangChain, a framework for building applications based on machine learning language models.
- Nomination 'Supply Chain Sentinel': Alexander Brandes (@notmyfault), maintainer of the Jenkins project.
- Nomination 'Audience Choice': Astro, a static site generator.
- Nomination 'OSPO Leadership' (Open Source Program Offices): Mercedes Benz, which actively uses open source software and published a corresponding manifesto.
- Nomination 'DevOps Champion': Katie Peters from TELUS.
- Nomination 'GitHub for Good': Ersilia, an organization developing open source software for medicine.
Additionally, GitHub introduced new application security testing tools that use machine learning and announced the launch in December of a chat with an AI consultant (GitHub Copilot Chat) to assist with code writing, error identification, and test creation, as well as answering development-related questions posed in natural language.
Source: opennet.ru
