Reference counting garbage collection

[Paul Rubin]

"Alex Martelli" <aleax at aleax.it> writes:
> And programmers used to reference-count semantics (finalization
> doesn't occur given reference loops, but without them it occurs
> instantly as soon as an object is found to be unreferenced) have
> proven to be very reluctant to abandon those semantics for m&s
> semantics (whether _implemented_ by m&s, copying, generational
> mixed-strategies, etc, isn't the point:-).

I think the whole idea of finalization is somewhat in conflict with
the traditional concept of GC, which is that GC simulates having infinite
memory by reclaiming objects that are no longer reachable.  Under that
model, nothing ever really gets finalized.  

Finalization is useful, to be sure.  I'm not enough of a language
connoiseur to know how it's been done in other systems except C++
and Java, which are both monstrosities (well, C++ anyway).  I may
dig out my CADR manual to see how it was done in the good old days :).