RE: viewpoints and multiple inheritance.
Chris Welty explains:
> No, it isn't. This is precisely the kind of confusion that (perhaps)
> derives from the Object-Oriented "code-sharing" paradigm (to
> paraphrase John Thompson in this list) that I would like to avoid.
> Perhaps these factors can lead one to consider arranging properties
> in a taxonomy, but once you do it you have a universal assertion, not
> a sharing of factors. Regardless of whatever they have in common, if
> you say cat is-a pet then it means ALL cats are pets. THAT is what
> it means, and nothing else. Single or multiple, it doesn't matter.
If someone is taking notes of key points, this seems to be one, the difference
between the meaning of multiple inheritance when seen from a logical
perspective, as Chris describes, as compared to a code-sharing view. Both are
no doubt legitimate and useful in their respective domains. The trick, perhaps,
is to clearly identify what those domains are, what the precise meaning is in
each domain, and exactly when and why one is appropriate in one context, but
not the other, and vica versa.