I think that's more of an issue with C++. Usually developers stick to known patterns and subsets. Sure, you can write "clever" code that's more aproppriate for

I think that's more of an issue with C++. Usually developers stick to known patterns and subsets. Sure, you can write "clever" code that's more aproppriate for IOCCC than as system-critical code, but that's the case with most programming languages. There is such a thing as "bad code" and "good code". A lot of code linters can catch "code smells", and features can often be disabled at the compiler level. On the other hand, not having to resort to inline assembler or writing 500x the amount of code in some cases is why we have such features. Complex features are useful in some (complex) cases, but are not to be overused.