I have a related concern with the idea that the Bitcoin Core source code is the Bitcoin protocol specification.
Eventually, the C++ language will go the way of COBOL. Very few people will be able to interpret and maintain the specification in the not-too-distant future.
COBOL was never designed as a general purpose language. Part of the problem with C/C++ is that we are absolutely, irrevocably stuck with it because it is so integral in so many things. It means that newer and potentially better-designed languages have slim chances of succeeding because of C++'s dominance. It isn't going anywhere anytime soon. [...]
To add to that, C/C++ compilers are much better understood and tested than compilers of newer languages. If only by merit of being around for much longer. So while newer languages may seem better designed and more appropriate they still have a couple of years, maybe even decades left until they reach the level of trust and performance of C/C++.
I mean just look at Ada. That language is even older than C/C++ but there's a reason why
to this day it's still being used for military and avionic software development.