There is one feature of the Objective-C language that I really love: the method naming. Let me explain with an example.
Here is a nasty call to a C function from the Win32 API that has 12 arguments:
hwnd = CreateWindowEx(WS_EX_LAYERED,
TEXT("Hello"),
TEXT("World"),
WS_OVERLAPPEDWINDOW,
10,
10,
400,
400,
NULL,
NULL,
hInstance,
NULL);
Pick an argument, any argument. What does it do? You can probably guess a couple of them, but basically you’re forced to look up the documentation. Sure, 12 arguments is a bit excessive, but even three or for argument functions can be ambiguous. What if you’re trying to understand a function call that is being passed three number literals as arguments? Even if you know the function, you’ll probably have to look up the documentation just to remember the order of the arguments.
Now for the equivalent in Objective C:
hwnd = [SomeClass createWindowExWithExtentedStyle:WS_EX_LAYERED
className:TEXT("Hello")
windowName:TEXT("World")
style:WS_OVERLAPPEDWINDOW
x:10
y:10
width:400
height:400
parent:NULL
menu:NULL
instance:hInstance
param:NULL];
Aren’t the named arguments great?
At first it feels redundant when you’re writing method declarations because the method name segments are basically the same as the argument name. The method declaration for the above method would look like this:
+(HWND) createWindowExWithExtentedStyle:(DWORD)extendedStyle
className:(LPCTSTR)className
windowName:(LPCTSTR)windowName
style:(DWORD)style
x:(int)x
y:(int)y
width:(int)width
height:(int)height
parent:(HWND)parent
menu:(HMENU)menu
instance:(HINSTANCE)instance
param:(LPVOID)param;
Yes, it is a bit redundant in the declaration, but reading code that calls the method is so much easier. Whenever I write something in another language, I miss my named arguments.