From: "Chr. Rippel" <chr_news@...>
Date: 2002-01-31T07:44:43+09:00
Subject: Re: strange behavior

"Paul Brannan" <paul@atdesk.com> wrote in 
....
> IMO, a method should not have to know the name of the class in which it
> is defined.  If I change the name of the class, then I must be very
> careful to make sure that I also change any code inside the class
> definition that uses the original name.  Additionally, if I
> cut-and-paste a function from one class to another, then I must also be
> careful.
> 
> In theory, changing a class name isn't (or shouldn't) be done often, and
> I should really consider why I am using cut-and-paste anyway instead of
> designing my code properly in the first place.  A wise person once
> said, though, that the diffence between theory and real life is real
> life.

One problem is that at the ``time of method body declaration''
you don't have access to the context - a.k.a binding.  It's a bit like
the use of Pythons ``self'' variable - if don't spezify ``self'' you 
cannot access self (in Python)

class A
        def A.meth; "a" end
        def foo(class_name = A)
            puts class_name.meth
        end
end
class  B < A
    def B.meth; "b" end
end
  
b= B.new
b.foo       # => a
b.foo(B) # =>  b


/Christoph