Bjarne Stroustrup's Plan for Bringing Safety to C++
https://thenewstack.io/bjarne-stroustrups-plan-for-bringing-safety-to-c/
Discussions: https://discu.eu/q/https://thenewstack.io/bjarne-stroustrups-plan-for-bringing-safety-to-c/
@c_discussions I really like this approach.
Herb Sutters "cpp2" is a really interesting idea. https://github.com/hsutter/cppfront
something that compiles to C++, i.e. high interop with existing c++ codebases, but an overall cleanup + more condusive to safety.
Retrofitting safety to C++ directly seems too convoluted to me.
The manageable complexity budget & syntax space is already spent.
I like Rust but accept C++ codebases need to continue, few can afford a clean start.
@c_discussions nice idea, but I'm afraid this is too little, too late and also too complex than needed. Also the proposal depends on C++ modules specified in '20 while in '23 we still do not have C++ modules fully working. Just one (VS) compiler from 3 (GCC/Clang/VS) somewhat implement them. So let's count, it'll take 6-10 years by committee to standardize and additional 5 years by compilers to implement the feature. So we're looking into ~2040 when this will be available in latest compilers plus additional ~5 years before it's available in platform (supported/trusted) compilers (embedded world). Do you think that industry will wait for 20 years for C++ to deliver crucial safety feature?