@vaporeon_ to oversimplify a bit it's like a c-style cast that has a lot more restrictions,and is checked by the compiler, to protect the programmer from making silly mistakes
@wallhackio What kind of silly mistake? In my mental model, a cast is a way to tell the compiler "I know what I'm doing, don't complain and just consider this type as this other type", and now you're telling me there's a version of that with extra checks?
@wallhackio Somewhat relatedly, can you tell me why the C++ compiler is complaining about this very normal code?
#include <stdlib.h>
int main() {
char *str = malloc(123);
}
$ g++ test.cpp
test.cpp: In function 'int main()':
test.cpp:3:21: error: invalid conversion from 'void*' to 'char*' [-fpermissive]
3 | char *str = malloc(123);
| ~~~~~~^~~~~
| |
| void*
Sure, a void* is not a char*, but the void* is intended to be casted to some other type because you can't dereference a void*, and in C this code just works... Why did they choose to make my life harder?!
@vaporeon_ it was a very deliberate design choice when C++ was made in the 80s to not allow implicit void* conversions. It's one of the few very deliberate situations C++ fails to be a superset of C. You have to explicitly cast a void pointer to the exact type you want it to be
You can disagree with this choice but it was a choice made from basically the very beginning
@wallhackio Am I supposed to not be using malloc() in C++ and be using something else instead?
@vaporeon_ @wallhackio it’s a very abnormal thing to do, to my understanding
@vaporeon_ @aescling there's almost always some sort of standard library container class for anything you'd use malloc for. I'd much sooner default to std::string in your situation
If the container doesn't exist then you can use the new keyword. I don't know of a situation you would use malloc other than to interface with C libraries or legacy C++ code