From: Dave Thomas <Dave@...>
Date: 2001-08-16T21:42:42+09:00
Subject: [ruby-talk:19890] Re: Why not?: Assigning to self

"Pit Capitain" <pit@capitain.de> writes:

> Why not? In this class hierarchy each square is a (specialized) 
> rectangle, so class Square also has a method #set_shape. If 
> called with width != height, the Square instance would "become" a 
> Rectangle instance. In a drawing program showing a square, why 
> shouldn't it make sense to be able to drag one side or one hotspot 
> in order to change the square into a rectangle?

If you were to allow that, what is the difference between a Square and
a Rectangle as classes? Why differentiate? Perhaps the underlying
Shape class should have a #has_equal_sides? method?


I can see a use for some kind of delegation-on-steroids facility,
where you could say:

   my_object.responds_as other_object

Implemented in the interpreter, this would give you useful semantics
without (much of) a speed penalty.

However, I can't see any use for 'becomes.' My bias here comes from
the fact that in Ruby classes are not types, and types are not
classes. (In a way it's a shame that there's a built-in #type method,
because this is misleading.) If we had a #becomes that swapped an
object's class, it would be a subtle but profound shift in the
underlying language, giving classes more significance than they
deserve.

Dave