What is the difference between pointers in C?

Question

C++ / C#

Greetings, I can’t figure out on my own what is the difference between the two methods of use, for example:
obj.show_nameand obj->show_name
Let's say we have a data structure:

typedef struct users
{ 
  char* name;
  char* password;
}user_t;

Next, we start working with this structure:

user_t mt1;
user_t *mt2;

int main()
{
  mt1.name = "Andrey";
  mt1.password = "Password";

  mt2 = malloc(sizeof(user_t));
  mt2->name = strdup("Andrey");
  mt2->password = strdup("Password");

  mt1.name = NULL;
  mt2.password = NULL;

  free(mt2->name);
  free(mt2->password);

  return 0;
}

When should mt1 be used, and when should mt2 be used? In my opinion, there is no difference in their work, while the implementation of mt1 is clearly shorter than mt2, based on this fact, I almost always use the mt1 method and do not understand why to use mt2.
I tried to find an answer in the literature (S. Prata), the theory is clear, but in practice I cannot understand the fundamental differences between these two approaches.

Answer 1

There is a concept of a stack. The stack is small: it is better not to allocate large objects on it. Therefore, a heap is also needed - memory is allocated there using malloc.
In addition, the lifetime of objects differs in the first and second cases. An object allocated on the stack lives until the end of the scope (usually until the closing curly brace). An object allocated on the heap lives until it is explicitly deleted (using the free method you forgot to call). Those. such dynamically allocated objects are also handy if you need to pass the object itself and/or manage its lifetime to another part of the program.
Well, if you want to modify any object (no matter how created) in some third-party function, then you need to pass a pointer to this object. As far as I remember, there are no references in C.
PS
For C, read K&R better.

Answer 2

There is also the symbol & which allows you to describe variables, access to which is syntactically no different from direct access, but in fact it is a link:

class MyClass
{
public int value;
MyClass(int _value):value(_value){}
}

MyClass *tmp=new MyClass(10);
MyClass &object=*tmp;
object.value=20;
delete(tmp);

99% of the differences are in the way we initialize (and deallocate), and in the last 10 years c++ has made so many changes around this.
The reason is an attempt to gain more control over the object's memory so that the compiler has more options to deallocate it.
ps , use & to define the types of parameters passed to functions so that it is not copied, but passed by reference, and at the same time, syntactically, you would work with objects directly.

Answer 3

Everything is simple.
Try how to dynamically allocate memory using only mt1 :) (Hint: it won’t work. You can still allocate memory through malloc using sizeof struct, but assigning the address of the beginning of the area to the mt1 variable will work)
Structures are an awesome tool for scheduling memory, especially when reading binary files with a fixed structure inside - counted N bytes into the area, imposed a structure on this area - and that's it, you can access the fields. That, actually in an example at you also becomes.