China’s open source with 7.55 million developers, what is the progress?

Author | Peng Huizhong Editor in Charge | Tu Min

Produced | CSDN (ID: CSDNnews)

It has been more than 20 years since the seeds of open source have been sown in the land of China, and now is the time when a hundred flowers are blooming and competing for beauty.

From the perspective of developers, according to GitHub 2021 statistics, the number of Chinese developers on GitHub has increased by nearly 1.03 million compared with the previous year, and now it has reached 7.55 million open source developers, ranking second in the world.

From an industry perspective, whether it is a traditional industry or an emerging industry, the penetration rate of open source is gradually deepening. So, what is the development of open source in various industries? What is the biggest stumbling block on the road to open source? What challenges and opportunities does enterprise open source face?

This issue of “Open Source Roundtable” starts with “How to promote the growth of various industries with open source in China?”

As the theme, Cao Hengkang, Global Vice President and President of Greater China of Red Hat, Li Wei, Deputy Director of Cloud University of China Academy of Information and Communications Technology, Li Yang, Co-founder & CTO of Kyligence, and Wang Wang, Vice Chairman of China Artificial Intelligence Open Source Software Development Alliance Dr., under the auspices of Jiang Tao, founder & chairman of CSDN and founding partner of Geekbang Venture Capital, jointly explore the true face of open source in industry applications.

Scan the code to view the playback

China’s open source is accelerating integration with the industry

Jiang Tao: In recent years, the penetration rate of open source in various industries has gradually deepened. Which industries are developing the fastest? Please share your experience.

Cao Hengkang:

In the report “The State of Open Source in Global Enterprises 2022” published by Red Hat, we interviewed nearly 1,300 enterprises around the world, and the results show that the acceptance of open source and the proportion of open source use by these enterprises are increasing year by year, and the growth rate is very rapid. 90% of the surveyed enterprises have basically regarded open source as their main IT development direction. The country has also ushered in explosive growth in recent years. Undoubtedly, this is related to the inclusion of open source in the “14th Five-Year Plan” in China and the use of open source as the strategic development direction of the entire country, and it has a lot to do with the development of technology. .

Open source is widely used in all walks of life. The above report is only for enterprise-level open source. If community open source is added, then its penetration rate in various industries is very high. In terms of industries, in the fields of finance, telecommunications, and automobile manufacturing, open source has become the main technology for its operation.

Li Yang:

The big data industry has also ushered in explosive growth in open source. When we first started doing open source, there seemed to be no competition in the industry. However, until recently, the open source big data track has been very crowded. It only took two years to change from the blue ocean to the red ocean. This is a great common ground.

Li Wei:

First of all, from the data point of view, a total of 72% of the Fortune 50 companies in 2020 are using GitHub’s platform to host code. Over 80% of domestic industry customers have used open source technology in software development and production.

Secondly, from the perspective of work contacts, from 2012 to 2017, the Internet industry was in the ascendant. The products and services of the Internet industry are based on open source, and Internet companies gradually open source their own projects from using foreign open source projects. Since 2017, the financial industry has begun to pay attention to open source, and in 2018, a community of financial open source technology applications, FINOC, was established. At that time, it was jointly initiated by us and Shanghai Pudong Development Bank, and now there are nearly 100 financial enterprises in it. Last year, the telecom industry established a corresponding telecom open source community. The recent rapid development is the automotive technology manufacturing industry, and an open source community has just been established this year.

Dr. Wang:

At present, the AI ​​industry is also at the forefront in open source. From the underlying framework to the upper-level model to the application layer, all parts of the entire AI industry chain are developing rapidly.

Jiang Tao: With the deepening of digitization, open source is more and more widely adopted. What is the biggest obstacle you encounter when developing open source in these industries?

Cao Hengkang:

Overall, from a business perspective, I think there are three barriers:

First, there is a lack of human resources. At present, the talent resources related to open source are insufficient, so many situations are often encountered. For example, when working on open source projects, it is difficult to make concerted efforts to open source some applications, but the open source talents of these projects are poached in a short time. To stop.

Second, talent and technology are not enough. Existing talents still need to further develop in the technical field.

Third, enterprises do not have a deep understanding of open source. Many companies regard open source only as a technology, but open source is not only that, it is also an ecology and a culture. The reason why open source has grown rapidly over the past 30 years is that its community culture is very strong, so enterprise IT staff should be encouraged to participate in the community, not just use open source technology.

Li Yang:

From the perspective of companies in the big data industry, I think there are two obstacles:

The first is the cost issue. Whether an enterprise uses its own technical personnel to cover the cost of using open source software, or purchases enterprise-level open source software to obtain stable and reliable services, the enterprise will consider which cost is lower.

The second is the difficulty of choosing. There are many open source big data analysis technologies on the market to choose from. In addition, each technology may have an open source version and an enterprise version, so it takes a lot of effort to choose the corresponding technology.

Li Wei:

From an industry perspective, I think there are three points of resistance:

First, there are barriers between traditional industries and open source. Most practitioners in the Internet industry are related to software and have a deeper understanding and mastery of open source. However, in traditional industries such as finance, it is challenging to integrate open source software into the compliance of internal development in the financial industry. Practitioners in the financial industry do not have enough technical reserves related to open source, so when using open source, they may need the support of third-party services to help them land, and at the same time, they need to improve the internal management system to ensure the compliance of open source use.

Second, the iteration speed of open source is too fast. Especially in recent years, there is a phenomenon that an open source technology is not yet familiar, and it seems that the next generation of alternatives has emerged, thus causing the cost of choice that Mr. Li said.

Third, open source security issues need to be resolved urgently. The iteration speed of open source is very fast, and there will inevitably be many loopholes, which requires enterprises to have certain technical capabilities to track in real time and avoid these problems.

Dr. Wang:

In the AI ​​open source industry, my two most noteworthy points:

The first is stability: AI models iterate very quickly, and there are often “technological revolution”-style subversions. Various big manufacturers are racing to launch similar models, which is dazzling.

The second is security: not only the security of AI itself, but also the vulnerabilities existing in open source, and the problem that code may be attacked after open source.

Jiang Tao: Today, digitalization penetrates into every industry.
At present, smart cars should be the next big digital development market, so how do you think open source will develop in this industry?

Cao Hengkang: The automotive industry is like another IT industry, it just integrates a complex operating system and many other IT technologies into a car.

Taking BMW as an example, it has always attached great importance to autonomous driving, and established an autonomous driving project many years ago, and also placed open source in a very core position in the autonomous driving project. The reason is the use of open source development methods. Very agile and fast.

In the process of autonomous driving, the entire system needs to process all information very quickly, so this industry has very high requirements on IT technology, including big data, artificial intelligence and the most popular edge computing (Edge computing) and other technologies, so intelligent The development methods of the automotive and Internet industries are similar, and they prefer to use open source agile development. At the same time, the automotive field also needs stability and safety, because autonomous driving is closely related to human life, so its standards are very high.

Li Wei:

Automotive is an industry where OT and IT converge.

On the one hand, there is currently no unified standard for the requirements of smart car operating systems and related software. The open source features make it difficult to unify this standard, so there are various solutions on the market. In this process, everyone needs to be compatible with each other and reach a consensus on a standardized solution, and open source is an important way to reach this consensus.

On the other hand, the automobile industry is not only its own construction, it is also closely related to the Internet of Vehicles and large transportation sectors. As far as we know, many auto tech companies are exporting software for smart transportation. Therefore, these open source software are also inseparable from the integration process of hardware vehicles and smart transportation.

Li Yang:

The car is a new big carrier. If the Internet was a big outlet in the first two years, then the next generation of advanced intelligent platforms like automobiles is a new channel for technology to realize commercial value.

I want to ask a question: Can you imagine that hardware is also open source? Is it possible that a whole car from hardware design to software architecture is all open source. If there is such a car, will anyone pay for it?

I guess most likely not everyone will. why?

Because end consumers don’t pay for a feature. There are many alternative functional solutions, but what really makes you decide to pay is the security, stability and reliability of the system. The end consumer pays for these non-functional parts.

Therefore, in terms of open source, manufacturers that can solve non-functional problems at new heights will bring greater value to the market and will also have greater profit margins.

How China Open Source Improves Talent Cultivation

Jiang Tao: Regarding the problem of lack of talent resources in open source, do you have any solutions to this problem?

Li Wei:

At present, many technology companies and foundations will have their own talent training and certification in some open source projects. In addition, some third-party organizations will also provide training services for open source talents.

On the other hand, our China Academy of Information and Communications Technology is also conducting school-enterprise cooperation. If you can start with software as a student, it’s great to think of open source as a very skilled skill. Therefore, it is very important to carry out basic talent training in colleges and universities, strengthen the understanding of open source, and convey the concept of open source.

Dr. Wang:

To cultivate talents, I think it is nothing more than starting from the following aspects:

First of all, it is necessary to increase the popularization of open source awareness, implant the awareness of open source into the training of talents in IT-related majors, and let everyone actively participate in open source projects while in college.

Second, in the process of corporate recruitment, candidates with high open source contributions are preferred. Interviewees should show their Homepage on open source platforms like GitHub, for example, to demonstrate their involvement in open source.

Furthermore, enterprises need to work hard to build an open source ecosystem. Just as Baidu, Huawei and some start-up companies will open source their own projects, companies have the responsibility to participate in the construction of an open source ecosystem. Only when the ecological environment is good can talents survive in this “fish pond”.

Cao Hengkang:

As a leader in open source technology, Red Hat is not only committed to developing software and serving customers, but also includes the following talent development initiatives.

First, every year, Red Hat sends people to various universities to promote open source technology and open source culture, so that more students can understand open source.

Second, Red Hat will hold an open source challenge every year, covering hundreds of colleges and universities across the country and thousands of students. We will establish a talent pool of award-winning students to help companies provide talents.

Third, Red Hat, in cooperation with TWT, has held the Container Cloud Competition for two consecutive years, allowing more enterprises to participate in the development of open source technologies. For two consecutive years, there are about 30,000 people and thousands of enterprises participating. In general, Red Hat provides a lot of talent development.

Jiang Tao:

The talent problem has indeed come at an urgent time to be solved. At present, the China Computer Federation has established the Open Source Development Committee, led by Academician Wang Huaimin, and is already doing open source open courses, starting with college students. The project is currently being promoted in Peking University and Tsinghua University, and CSDN and Beihang are also doing it, and the response has been very enthusiastic.

In addition, the Institute of Software of the Chinese Academy of Sciences has also held the “Open Source Software Supply Chain Lighting Program” program. I think China’s open source has taken action in the education sector. I believe that with the injection of the next generation of talents, their understanding of open source will be better than our generation. The deeper it is, the greater the contribution will be.

How to crack open source problems in enterprise use

Jiang Tao: In the process of enterprises using open source, they will also encounter many concerns and problems. How do you think we should solve such problems?

Li Yang:

In fact, from the perspective of enterprise open source suppliers, we can help enterprises solve problems by providing enterprise-level open source software and services. For example, on the basis of facing a variety of open source technologies, we can provide enterprises with selection suggestions. First of all, we can clarify what is wrong with the enterprise, and then we can help the enterprise to choose a suitable technology and provide safe, reliable and stable services beyond functions. A business doesn’t necessarily need employees to troubleshoot all technical issues. If the positioning of the enterprise is clear, it is not difficult to provide open source technologies and services to the enterprise.

Cao Hengkang:

The real core competitiveness of enterprise open source is stability, reliability and security. Without such a guarantee, the shutdown of the banking system, the failure of the self-driving car, and the collapse of the telecommunications system would have serious consequences. Therefore, when enterprises choose open source technology, they should first clarify their positioning and consider whether to choose community open source or enterprise open source.

The benefits of community open source are a good ecology and rich innovation. You can learn and try cutting-edge technologies in community open source. But when you use open source in production, reliability, stability, and security are the first priorities, so enterprise open source is more suitable.

However, even if customers choose the so-called enterprise open source, the market is still mixed. Many so-called enterprise open source software, which may be under the banner of open source, but do not develop software in accordance with the principles of open source, will become a security loophole , turning the so-called open source into closed source, so when enterprises choose open source, they should be careful not to step on the thunder.

How to Govern the Security Conundrum of Open Source

Jiang Tao: Open source security has always been a problem that everyone is very concerned about. What kind of solutions are there now?

Li Wei:

Open source security issues can generally be classified into three categories.

First, in terms of vulnerabilities, some open-source technology companies provide open-source security tools and software that can scan for open-source vulnerabilities.

The second is licensing-related issues. Since many open-source projects are nested with other open-source projects, there may be internal license conflicts. At present, there are corresponding software for the analysis of code formation and the analysis of whether the corresponding license conflicts after formation.

The third is the security problem in the supply chain, which is relatively difficult to find. For example, some security problems occurred upstream of open source software distributions. In this case, the security control of software sources is very important, and it is necessary to rely on Party A’s own security management capabilities.

Li Yang:

From the perspective of serving customers, our company’s slogan is: stability first, safety first. This is also the core value of our enterprise-level open source services. Every new security vulnerability recently is a first-level red alert within the company. The entire production and research team will mobilize all forces to solve the problem as soon as possible, and inform customers about this security vulnerability and your current production environment as soon as possible. Does it matter.

If it has an immediate impact, we will help users block the “flood” in the first place;

If there is no impact for the time being, we will also arrange time to fix it to ensure that it will not be affected in the future.

Cao Hengkang:

Judging from the enterprise open source report, the number one concern of the CEOs of nearly 1,300 IT industry leaders around the world is security. This is true whether it is for professional software or enterprise open source software. As the open source technology becomes more and more mature, people are more and more aware of open source. A data survey shows that 89% of IT executives believe that enterprise open source software is more secure than general professional software. The reason is the following five points:

First: the code of the open source software is in your own hands, and you can check the reliability of the source of the code at any time.

Second: Security patches are well documented, and enterprise open source software suppliers will have dedicated personnel to trace patch vulnerabilities to upstream through scanning.

Third: Suppliers can provide enterprise open source vulnerability patches in a timely manner.

Fourth: Enterprise open source software vendors will help you review intellectual property related issues to ensure that the code is OK.

Fifth: The supplier will provide training to customers and inform them of the points that need to be paid attention to during use.

It is because of the above points that many companies believe that using enterprise open source software is more secure.

Advice for open source practitioners

Jiang Tao: What advice do you have for open source practitioners?

Li Wei:

First of all, the current consensus in the industry is to embrace open source. The role of open source in promoting the development of different industries needs to be affirmed by everyone.

Second is the normative issue, and security is probably the most important thing in open source.
In addition, in the process of use, the technical path also needs to be carefully selected.

Therefore, we need to walk on two legs of “development” and “norm”.

Li Yang:

If someone wants to join the wave of open source entrepreneurship, my suggestion to everyone is: Entrepreneurs need to clarify their own value in the entire open source software ecosystem, and the paid value you create should usually not be the development of a new function for users. The value point is how to ensure that the solutions provided are safe, stable and reliable.

Cao Hengkang:

First of all, it is very important to understand the open source culture and philosophy. I encourage enterprises to understand the organizational structure of open source, integrate the spirit of open source into the enterprise, pay attention to participation and contribution, and truly achieve “I am for everyone, and everyone is for me”.

Secondly, for enterprises, in order to obtain stable and reliable guarantees, it is recommended to use enterprise-level open source. What enterprises buy is not technology, but services and guarantees. Open source does not mean free, and as an enterprise manager, it is necessary to pay a relatively reasonable price to obtain reliable protection.

Dr. Wang:

To quote a buzzword in the field of AI: language learning is learning of learning. That is, language learning is not something that can be learned in a short period of time. There is a threshold for language learning of open source software, so I also hope that more companies, entrepreneurs and even individuals will continue to accumulate experience in open source.

That’s all for this issue of Open Source Roundtable. With the great development of China’s digital economy, the penetration of open source technology in various industries will become deeper and deeper. At the same time, there are still many challenges, including talent issues and open source professionalism. Therefore, we need more people to become open source contributors and participate, and hope that more people will pay attention to our follow-up “Open Source Roundtable” series of interviews.

The text and pictures in this article are from Sina.com