Re: Idempotence and "Replication Insensitivity" are equivalent ?

vc wrote:
> Marshall wrote:
> > Phil Carmody wrote:
> > > "Marshall" <marshall.spight_at_gmail.com> writes:
> > > > Consider min(). Min over unsigned 32 bit ints has an
> > > > identity: 2^32. However, min over arbirtrary precision
> > > > ints does not have an identity.
> > >
> > > If 2^32 is an identity over 32-bit ints, then +infinity is
> > > an identity over AP ints.
> >
> > I understand that viewpoint, however +infinity is not an
> > integer, so there are typing issues involved with that
> > idea. My sense is that those issues cause more
> > problems than they solve.
> >
> >
> > Marshall
>
> Neither is 2^32 an identity (wrt min) since it does not belong to the
> set of 2^32 unsigned integers which presumably contains zero.

D'oh! That's the second time I've made that exact fencepost error in a month!

Marshall Received on Thu Sep 21 2006 - 18:19:15 CEST