For a few months now C++ has had the biggest share of active contributors in open source projects tracked by Open Hub.
Each contributor is counted once for every language they have contributed in in the last month.
C++ has also most commits and code lines changed in the last month, but I think the number of active contributors most accurately reflects language popularity, that this method defines as willing to contribute in the language. Now one might protest that the project dictates the language to use; fair enough, start a project in Malbolge and see how many contributors you get. Moreover C++ comes only fifth in the number of active projects, after Python, shell script, JavaScript and C. Clearly C++ projects have more active contributors on average than those languages.
I'm glad to see it because despite Rusts hype and popularity on HN I love to write C++ and I am even starting new projects in it occasionally. I have read many times that I should be starting new projects in Rust these days. The last project I started was a 128-bit assembler for RISC-V of all things, in modern C++. I understand that certain languages are better suited to certain things, however I feel like I am solving problems in C++ the way that I want to, I guess, in a very opinionated way. And I don't like to compromise on the design. It's not even about performance, even though C++ is very clearly the king.
That said, I am happy that we have a plurality of languages to choose from, and that they are all popular, relative to each other.
> I feel like I am solving problems in C++ the way that I want to, I guess, in a very opinionated way. And I don't like to compromise on the design.
Me too! I'm glad that Rust has managed to champion extensive static analysis forward, but at the same time it's clear that rustc can prove only a proper subset of correct programs correct. I don't like the idea of being restricted to a subset of correct programs. The philosophy of C++ is to allow anything that could be correct, but you can still use tools to detect suspicious code. The difference is that you can ignore the false positives.
3 comments
[ 3.0 ms ] story [ 23.6 ms ] threadC++ has also most commits and code lines changed in the last month, but I think the number of active contributors most accurately reflects language popularity, that this method defines as willing to contribute in the language. Now one might protest that the project dictates the language to use; fair enough, start a project in Malbolge and see how many contributors you get. Moreover C++ comes only fifth in the number of active projects, after Python, shell script, JavaScript and C. Clearly C++ projects have more active contributors on average than those languages.
That said, I am happy that we have a plurality of languages to choose from, and that they are all popular, relative to each other.
Me too! I'm glad that Rust has managed to champion extensive static analysis forward, but at the same time it's clear that rustc can prove only a proper subset of correct programs correct. I don't like the idea of being restricted to a subset of correct programs. The philosophy of C++ is to allow anything that could be correct, but you can still use tools to detect suspicious code. The difference is that you can ignore the false positives.