Can China Build a World‑Class Operating System? Lessons from a DIY OS Journey

Feasibility of a Commercial‑grade Operating System

Creating an operating system that can compete with Windows, macOS, Linux, or Android requires more than a few thousand lines of code. A modern OS typically contains tens of millions of source lines, a mature driver ecosystem, extensive user‑space libraries, graphical subsystems, security frameworks, and a global developer community. Building such a platform demands massive human, material, and financial resources, as well as long‑term coordination among hardware vendors, software developers, and application providers.

Core Technical Requirements

CPU Architecture Support : Understanding instruction sets (e.g., x86, ARM), privilege levels, and boot procedures.

Memory Management : Physical and virtual memory allocation, paging, segmentation, and protection mechanisms.

Process and Thread Scheduling : Context switching, priority handling, and inter‑process communication.

Device Drivers : Low‑level I/O handling for storage, networking, graphics, and peripheral devices.

File Systems : Reliable storage abstractions, journaling, and permission models.

Security Subsystem : Authentication, access control, isolation, and update mechanisms.

User‑Space Libraries and Runtime : C standard library, POSIX APIs, and language runtimes.

Graphical User Interface : Windowing system, compositing, and input handling.

Each of these layers must be integrated, tested, and continuously maintained, which explains why existing operating systems have evolved over decades with contributions from thousands of developers.

Case Study: Building a Tiny Kernel (LycheeOS)

The author’s personal project demonstrates how a minimal kernel can be constructed from scratch. The development workflow included:

Studying IBM BIOS interrupt manuals to understand real‑mode boot services.

Reading the Intel CPU architecture manuals for protected‑mode entry, segmentation, and paging.

Analyzing source code of early Unix‑like systems such as Minix and Linux v0.12 to learn kernel organization, system calls, and driver models.

Implementing a bootloader that switches the CPU from real mode to protected mode.

Writing basic io routines for keyboard and screen output using BIOS interrupts.

Developing a simple memory manager that tracks free frames and sets up page tables.

Creating a rudimentary scheduler that performs round‑robin context switches between two demo processes.

Testing the kernel on an emulator (e.g., QEMU or Bochs) to verify bootability and basic task switching.

The resulting kernel, although limited to a few hundred lines, can run on an emulator and serves as a proof‑of‑concept for low‑level system programming. The source code is publicly available at:

https://github.com/ahhuiyang/LycheeOS

Institutional Barriers in China

Both enterprises and research institutes face systemic obstacles:

Enterprises prioritize short‑term profitability and are reluctant to invest multi‑year, high‑risk OS projects without clear market returns.

Research institutes often focus on high‑profile, theory‑driven work and lack incentives to develop large‑scale software infrastructure.

Universities may have knowledgeable students but frequently lack sustained mentorship, dedicated hardware resources (e.g., fixed IP servers), and a culture of preserving and evolving student code contributions.

These factors limit the formation of the collaborative ecosystems that historically produced operating systems such as Minix, Linux, and BSD.

Recommendations for Building a Domestic OS Ecosystem

To overcome the current limitations, the author suggests:

Introduce programming and systems‑level education at the primary and secondary school levels to develop a larger base of skilled developers.

Leverage university curricula to integrate an open‑source OS kernel project into operating‑system courses, allowing students to contribute to a shared codebase.

Establish long‑term institutional support (e.g., dedicated servers, fixed IP addresses, version‑control infrastructure) to enable continuous development and maintenance.

Encourage open‑source collaboration across universities and industry to gradually grow a compatible driver and application ecosystem.

With sustained investment in education, infrastructure, and community‑driven development, a future generation of Chinese engineers could create operating‑system prototypes that may eventually evolve into commercially viable products.