Question

How bad is it to use a built-in object, type or whatever as something it’s blatantly not, even if it performs in exactly the way you need? Assuming it’s clearly commented what you’re doing.

For example, to use a single CGPoint to represent a start and end point on a single line. Should I really be defining my own identical but differently named struct instead?

Edit: OK, using…

struct FloatRange {
    CGFloat min;
    CGFloat max;
};
typedef struct FloatRange FloatRange;

…instead. Are there no exceptions to the rule?

Answer 1

For type safety, you should be defining your own.

Also remember you can’t do someRange.start or someRange.end. Rather, it would be the inscrutable someRange.x or someRange.y.