Question

Though the question is basic, but I found it very crucial to understand to proceed with IOS programming. Sometimes we used to declare only instance variable and we don’t set any associated property to it. Some where we just declare the properties and use synthesize to get or set the values. Sometimes I feel necessary to declare both in code, when compilation gives me warnings! What’s the fundamental behind the manipulation of properties in Objective C. I know the basic requirement to create getter and setter for any instance variable, but when?
I have seen it many times we don’t use property at all, and after that too we easily set and get variable’s value. Also, different types of property like atomic, non atomic, strong, retain are very unclear to me. XCODE up-gradation to 4.2 has shaken my concepts about memory management. Can somebody please clear the cloud over my mind?

Answer 1

Properties are always the preferred way over direct ivar access, mostly of following reasons:

• You can override a getter or setter in a subclass
• You can define the "assigning behavior" (namely copy, assign, retain/strong, weak)
• You can synchronize ivar access

Keywords:

• copy: The object is copied to the ivar when set
• assign: The object’s pointer is assigned to the ivar when set
• retain/strong: The object is retained on set
• weak: In ARC this is like assign, but will be set to nil automatically when the instance is deallocated, also used in a garbage collected environment.
• nonatomic: The accessor is not @synchronized (threadsafe), and therefore faster
• atomic: The accessor is @synchronized (threadsafe), and therefore slower

Generally you should always synthesize an ivar. If you need faster access for performance reasons you can always access the synthesized ivar directly too.