Re: dip Notions 2 Major Errors

From: Mark Nicholls (Nicholls.Mark_at_mtvne.com)
Date: 09/17/04 

Date: 16 Sep 2004 23:12:05 -0700

This was slightly wrong, I apologise for any confusion.....(confusion!...surely not)

>
> i.e. if I can draw this.
>
> [C]-------->[A]
>
> goes to
>
>
> [D]------>[A']
> | |
> | |
> v v
> [C']------->[IA]
>
> we can constuct D to be substitutable *class* (instance of metaclass) for the old C.
>
> class C is not the same as class C', they are not even substituble classes.
>
> this is the effect of the application of subtyping to a simple
> client/implementation interfaction.
>
> I see no inversion.
>
> The substituable class D (by Liskov, instance of metaclass), now has 2 dependencies going from it where
> the previous 'version' had one.
>
> So even the arrow had **not** been turned around, as some would claim as an
> inversion.
>
> QED.