What's the meaning of static variables in an implementation of an interface?

Question

I don't quite understand static variables when defined in the implementation of an interface. In methods I do understand how they differ from local variables, but not when defined directly in an implementation.

Look at these examples. What difference do these two make practically?

#include "MyClass.h"

@implementation MyClass
int myInt;
...
@end

And:

#include "MyClass.h"

@implementation MyClass
static int myInt;
...
@end

myInt is in both cases visible to all the methods, and if I interpreted a test I ran correctly, myInt will in both cases be the same variable for different instances of the class.

Answer 1

Unfortunately, it has different effects depending on where you use it.

Static Functions:
By default, all functions have a global scope. The static specifier lets you limit the function’s scope to the current file.

Static Local Variables:
When you use the static modifier on a local variable, the function “remembers” its value across invocations. For example, the currentCount variable in the following snippet never gets reset, so instead of storing the count in a variable inside of main(), we can let countByTwo() do the recording for us.

// main.m
#import <Foundation/Foundation.h>

int countByTwo() {
    static int currentCount = 0;
    currentCount += 2;
    return currentCount;
}

int main(int argc, const char * argv[]) {
    @autoreleasepool {
        NSLog(@"%d", countByTwo());    // 2
        NSLog(@"%d", countByTwo());    // 4
        NSLog(@"%d", countByTwo());    // 6
    }
    return 0;
}

This use of the static keyword does not affect the scope of local variables.
Read more about the static keyword.