Py2K wishes

[Paul Prescod]

Python 2 isn't under development yet. I didn't expect you guys to try so
hard to make my proposal formal. I'll be more careful with my ramblings
in the future.

Gordon McMillan wrote:
> 
> Your implications don't make sense to me, either. Does the
> first assert imply that MI is passe? 

Good point. It should be __fallbacks__. That makes it more useful
anyhow.

> Does the second imply
> that the difference between bound and unbound methods (and
> the reasons for the difference) have disappeared?

I admit that bound/unbound methods are one of the areas of Python that I
have to constantly re-learn because it doesn't naturally stick in my
brain. You are also right that I can't easily map that into __fallback__
because instances do getattr magic to methods that they do not to other
attributes. _fallback__ would be fine for the other attributes but
that's half a loaf at best.

I'm not 100% sure why Python is specified in terms of runtime re-binding
instead of "vtables" but it does cause somewhat of a problem here. Since
I am talking about Python 2, I can dream that the binding stuff is
simpler there. :)

 Paul Prescod