[chihuahua]

The way C++ has developed over the past 20 years seems similar to someone starting with an algorithm that fails for some edge cases, and patching the behavior with a different hack for each edge case, which breaks other cases, then patching those, and on and on forever.

[ryandrake]

I think the way to be successful with C++ is to 1. Pick a sensible subset of the language that you allow in your project, and ban everything else. How much that subset should include is a valid debate and reasonable people can disagree, but I don't know of any successful C++ project that just YOLOs every part of the language into the project. And 2. (related) Pick the earliest possible standard that your team can live with, and don't give in to the temptation of cherry-picking anything from a future standard. For instance, the decision of switching from C++14 to C++17 should be a major debate full of fistfighting.

[okanat]

Things start to break apart when you have dependencies that adopt newer standards or use broader features. There is only so much you can do unless you would like to reimplement libraries like SKIA, doctest, Qt6 or any modern game engine. It gets worse with security and updates. At some point a library will require a newer standard otherwise you have to adopt the entire codebase and assume the entire responsibility of all security updates.

At that point you are slowly rewriting the universe. So you can also do it in Rust tbh (which provides seamless updates and integration between epochs/editions).