Why New Programming Languages Are No Longer Being Created

Checking the TIOBE programming‑language index shows that over the past 20 years the dominant languages have been C, Java, Python, C++, C# and JavaScript.

Case study: automotive‑electronics project (2015)

During a technical‑selection meeting the lead chose C for low‑level drivers, C++ for the application layer and Python for scripting, and the decision was made in under five minutes.

Impact of product lifespan

Embedded systems in vehicles often have lifespans of ten years or more. Selecting a language that is only a few years old, with an immature compiler or ecosystem, creates a risk: if the language loses maintenance, the entire product would need a costly rewrite.

Embedded‑toolchain lock‑in

ARM’s CMSIS library, vendor SDKs and RTOS APIs are all written in C. Adopting a new language would require re‑implementing these components, dramatically increasing development effort and long‑term maintenance cost.

Failed Rust migration (2018)

A startup attempted to rewrite its embedded firmware in Rust to gain memory safety and performance. After three months the project was abandoned because essential components—CAN‑bus stacks, Modbus libraries and bootloader solutions—were either unavailable in Rust or of insufficient quality. The team reverted to C, where mature open‑source libraries were readily usable.

Hiring and team‑ramp‑up constraints

Job postings for “Kotlin Native expert” attract very few applicants, whereas “C/C++ expert” generates a large response. The hidden costs of training, recruitment and team coordination can outweigh any technical advantage of a newer language.

A technology manager of a 20‑person team reported that a half‑year learning curve for a new language would stall project momentum.

Ecosystem dominance

Java’s continued relevance is supported by the Spring ecosystem and numerous enterprise frameworks. Python’s popularity is driven by NumPy, TensorFlow and related scientific libraries. JavaScript dominates front‑end development through npm’s millions of packages.

Conditions for new‑language success

Go has established a foothold in cloud‑native environments, and Rust has gained traction in systems programming. Both solve problems that existing languages handle poorly and receive sustained investment from large companies or active open‑source communities. Languages lacking these factors tend to fade quickly.

Conclusion

The choice of programming language has become an ecosystem problem rather than a pure technical one. A smaller set of mature languages yields a healthier ecosystem and reduces the pressure on developers to constantly learn new syntaxes.

Code example

来源丨
经授权
转自 良许Linux（ID：liangxulinux）
作者丨
良许