Once you put deep learning in the same sentence as quantum computing, I can’t take the argument seriously any more. One has delivered state-of-the-art results in self-driving cars, machine translation, recommendation systems, and even robotics. The other has factored 21 into 3x7.
I would argue that Go and Rust are most likely to take hold on the low-level side, while data science takes hold of a new post-Python paradigm with static typing: maybe Julia, Jax/Flax. That much isn’t clear yet. I think raw JavaScript will continue to decline in use, possibly being overrun by WASM or an easily-packaged Electron-like environment in the cloud. Think about being able to run Python in a Docker container, that’s all run in a JavaScript environment.
For those that don't want to go to yet another medium.com article: Rust, Python, Ruby, Golang, GraphQL, Angular, and JavaScript, with an honorable mention for WebAssembly.
For an article aimed at programmers in general, Rust seems somewhat specialized to be in first place. Ruby seems even more out of place in third.
I started writing in C back in the late ‘70s. At the time BASIC, Fortran, and COBOL dominated, and Pascal was the supposed future, the Rust of its day. C still relevant today, the others not so much.
I learned SQL in the mid-80s, still relevant, I haven’t worked on a business application or web site that doesn’t use SQL in 20 years of web development.
I started working with PHP in 2000, when my employer ditched ASP to save money on licensing. More relevant today than ever, and I don’t expect it to go away because of the huge installed base and good fit to requirements.
The cool languages that will supposedly take over almost never do. Smalltalk, anyone? At various times in my career I heard that Modula 2, Ada, and even Scheme would dominate programming. Instead we got C++ and Java from out of left field, and Javascript now sits near the top of most-used languages.
Languages don’t “win” or even survive based on technical merits or programmer preference. They either solve problems and attract a community that leads to critical mass, or they don’t. I’ll be surprised if even 3 of the 7 languages listed in the article have widespread use in 5 years.
5 comments
[ 3.1 ms ] story [ 20.3 ms ] threadI would argue that Go and Rust are most likely to take hold on the low-level side, while data science takes hold of a new post-Python paradigm with static typing: maybe Julia, Jax/Flax. That much isn’t clear yet. I think raw JavaScript will continue to decline in use, possibly being overrun by WASM or an easily-packaged Electron-like environment in the cloud. Think about being able to run Python in a Docker container, that’s all run in a JavaScript environment.
For an article aimed at programmers in general, Rust seems somewhat specialized to be in first place. Ruby seems even more out of place in third.
I learned SQL in the mid-80s, still relevant, I haven’t worked on a business application or web site that doesn’t use SQL in 20 years of web development.
I started working with PHP in 2000, when my employer ditched ASP to save money on licensing. More relevant today than ever, and I don’t expect it to go away because of the huge installed base and good fit to requirements.
The cool languages that will supposedly take over almost never do. Smalltalk, anyone? At various times in my career I heard that Modula 2, Ada, and even Scheme would dominate programming. Instead we got C++ and Java from out of left field, and Javascript now sits near the top of most-used languages.
Languages don’t “win” or even survive based on technical merits or programmer preference. They either solve problems and attract a community that leads to critical mass, or they don’t. I’ll be surprised if even 3 of the 7 languages listed in the article have widespread use in 5 years.