| ▲ | j1elo 8 hours ago | |||||||||||||||||||||||||
A bit more if we consider the "bugfixing" release that was C++14 :) But yeah it makes sense, given how that was the jumpstart of the whole modernization of the language. I believe it was a big undertake that required the time it took. Still years have passed and footguns keep accumulating... it wouldn't hurt to have a mechanism to optionally drop the old cruft from the language. Otherwise everything stacks on top in the name of backwards compatibility, but at this pace, how will C++36 look like? | ||||||||||||||||||||||||||
| ▲ | mackeye 7 hours ago | parent | next [-] | |||||||||||||||||||||||||
a member of the c++ committee (herb sutter) is writing an compiler for an alternative c++ syntax [0], to c++, with the intent to restrict some semantics of the language for less UB, surprises, etc. i think less implementation-defined behavior is incredibly important; rvo vs std::move, dynamic function call optimization, i wish i didn't have to search asm to check for... | ||||||||||||||||||||||||||
| ▲ | pjmlp 7 hours ago | parent | prev [-] | |||||||||||||||||||||||||
While I share the sentiment, compare C#14 with C# 1.0, Java 25 with Java 1.0, Python 3.14 with Python 1.0. While C++ might be worse, when you have 300+ doing proposals every three years, others aren't safer from similar churn, even if on smaller volume, and trying to keep backwards compatibility going. And we all know what happened in Python. Also Rust editions contrary to what many think, only cover a specific set of language evolution scenarios, it isn't anything goes, nor there is support for binary libraries. As for a better C++, contrary to C, where it is business as usual, there are actually people trying to sort things out on WG21, even if isn't as well as we would like to. "Making C++ Safe, Healthy, and Efficient - CppCon 2025" | ||||||||||||||||||||||||||
| ||||||||||||||||||||||||||