David Farber wrote:  "Given that we have solution that is right, now we just
need to work on making it faster. Gee, isn't that the story of Smalltalk's
life?"

David, I will not bicker further on this issue today.  I agreed with
Martin's statements, including the one about IDENTITY being important.
Given that requirement, the solution that was proposed is in fact not right.

Also, I assumed that Ralph Johnson's speed tests were valid.  A general 5%
decrease in performance is more than significant.  Perhaps you have time to
write some tests to verify this claim.  I'm afraid I don't have time for
that today.

And finally, I believe it was also mentioned that a blind copy of the
receiver's collection could be very costly if the collection is very large.
I'd rather not blow the stack when I call removeAll:.

For those reasons, I stand by my earlier statement that we should not change
this method to what was proposed by Richard O.

---==> Chris