@secana good luck. If they are as angry as me about all the unfair and false claims about C++ being an unsafe language where people compare the latest and greatest #rust compiler against a 10 year old C++ compiler, thereby ignoring all efforts made in fixing defects of the language, you are not going to have a good times. #cplusplus
Nonetheless we have to start to work together and find a middle ground where we honestly can discuss strengths and weaknesses of both systems.
@markuswerle @secana I see no (deprecation) warnings when I write memory unsafe code with "old C++" and compile it with the latest and greatest compiler (gcc /clang).
So even IF what you might refer to as "modern C++" is memory safe, let's please keep things objective and not redefine common terms like "C++" as being "modern C++ where you have to pay attention yourself that you don't use any of the old C++ constructs which are part of the official language because otherwise it is a skill issue".
@mo8it @secana First, memory safety is a rather small subset of code safety. People from the #Rust community seem to think they have the holy grail in their hands because they *may *have solved one minor aspect - if they adhere to certain rules like "do not use keyword unsafe", which is not easy if you are talking to metal directly.
@markuswerle @mo8it @secana oh come on, that's ridiculous. Memory safety doesn't mean that it is possible to write safe code; that would be a useless definition, as all languages would be memory-safe by that definition.
Memory safety means that the language by design *guarantees* safety, i.e. that it is impossible to have UB and write unsafe code. C++ is definitely not memory-safe by that definition.
Rust is, and it is pretty unique in the sense that it has compile-time memory safety.