Smalltalk has a mechanism for "unde?ning" a method. Speci?cally, if a class A has method m, then a programmer may cancel m in subclass B by writing m:

self shouldNotImplement

With this declaration of m in subclass B, any invocation of m on a B object will result in a special error indicating that the method should not be used.

(a) What effect does this feature of Smalltalk have on the relationship between inheritance and subtyping?

(b) Suppose class A has methods m and n, and method m is canceled in subclass

B. Method n is inherited and not changed, but method n sends the message m to self. What do you think happens if a B object b is sent a message n? There are two possible outcomes. See if you can identify both, and explain which one you think the designers of Smalltalk would have chosen and why.