Question

Looking into the .NET code, Rectangle.Size returns new Size(Width, Height). Why did Microsoft choose this pattern? Personally, I would have thought that Size would be stored within the structure, and Rectangle.Width, for example, would return Size.Width. This would prevent a new structure from being created every property call. I’m guessing that there are some properties of immutability that influenced this decision, but I’m not sure what.

Answer 1

Size is a struct, so it’s not like it’s creating a new object on the heap. It will create a new copy of a Size value whatever you do.

I can’t see that it’s going to make much difference either way, to be honest. Given that the Width and Height properties of Size are inlined, I can see that there wouldn’t be much penalty from storing a Size as you suggest… but equally I can see that the constructor for Size is so trivial that the JIT may well be able to convert the Rectangle.Size property to almost exactly the same native code.

So I agree it’s a slightly odd decision, but I don’t think it’s going to hurt anyone much. Perhaps it makes serialization simpler or something like that.