Use of -> in Cocoa

Question

I'm used to create properties for pretty much everything. Probably too much, given. I also use instance variables, but I don't reference them with self->myvar but rather just myvar.

I recently was exposed to a code using a lot of self-> where I would have used a property. When I said one of the reasons I use @property (retain) is because I don't want to have to explicitly retain my object, I was told that I am "lazy". It's probably true, but it's also that I want to avoid bugs where I would have forgotten to retain.

Anyways, generally, what do you guys think about -> in Cocoa code?

Answer 1

Laziness is a virtue.

I use -> in copyWithZone:, to access the copy's instance variables. (I don't use property accessors here for the same reasons I don't use them in init or dealloc — they may trigger side effects that would be bad in the half-initialized copy.) I also use self-> in this context because I like the way that reads:

other->foo = [self->foo copy];
other->bar = [self->bar copy];
other->baz = self->baz; //Non-object or object not owned (may be omitted entirely in the latter case)

I can't think of another context in which I use -> on objects.

Answer 2

IMO there's not much value in using -> over plain ivar access. Sometimes it's useful for debugging purposes if you wanna access a different object's ivars. If it's just about differentiating ivars from local vars, you can use one of the established naming schemes (don't go into detail here) just as well.

The use of properties is a whole different story. IMO in most cases one should prefer using properties than plain ivars. Automatic retain is one reason, a good abstraction another.