Rust Foundation Releases Problem Statement on C++/Rust Interoperability (foundation.rust-lang.org)

submitted 1 month ago by snaggen@programming.dev to c/rust@programming.dev

7 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[-] onlinepersona@programming.dev -1 points 1 month ago* (last edited 1 month ago)

I bet this won't have an impact on memory safety and interop means C++ compilers have to be stricter about memory layout and reduce unspecified edge cases.

Anti Commercial-AI license

[-] robinm@fosstodon.org 0 points 1 month ago

@onlinepersona @snaggen Indirectly it can. Recent studies showed that old code is very unlikely to have security issue. This means that if all new code can be in Rust, while keeping the old code in C++ will be much more secure that rewrite all C++ (because by definition rewrite have more bugs since its new code). So interoperability is both safer and cheaper.

[-] 5C5C5C@programming.dev 1 points 1 month ago* (last edited 1 month ago)

It makes total sense that new C++ will contain a higher percentage of bugs than old C++, but after being an almost full time Rust dev for the last two years, you will not convince me that new Rust code has more bugs than old C++ code.

So far I have yet to come across a bug in any of my Rust code that made it into production. All issue reports from users are still related to the C++ code base that we haven't managed to fully divorce from.

The only advantage to C++ interop is that managers want to see new code get deployed immediately and continuously. They don't want to wait until the corporation's billions (literally) of lines of code are all rewritten in a new language before they start to see the benefits of that transition.

[-] robinm@fosstodon.org 1 points 1 month ago

@5C5C5C I found back the study I was talking about

https://www.theregister.com/2024/09/25/google_rust_safe_code_android/

> The good news for organizations with a lot of unsafe legacy code is that rewriting old code in new languages probably isn't necessary.

> That's not to say old bugs miraculously become unexploitable. Rather, the overall density of vulnerabilities diminishes – a statistical win but not a guarantee of safety.