Re: Entity and Identity
Date: Mon, 28 Sep 2009 11:22:52 -0700 (PDT)
On Sep 28, 12:12 am, Clifford Heath <no.s..._at_please.net> wrote:
> Brian wrote:
> > You are misrepresenting what I wrote. I did not assert that location
> > is part of state. I said that a difference in location constitutes a
> > difference in state. I said that because an object that can have
> > state can only have one state at a time and because that state can
> > only occupy one location at a time. It therefore follows that if
> > there is more than one location occupied at a given time, then there
> > is more than one object at that time.
> I see. Perhaps that's the issue that folk have with the O-O model,
> that there is no agreed method for identifying two objects as
> representing the same "thing". It's very hard to reason about a
> thing when you can't even identify it, so O-O is necessarily lacking
> the kind of theoretical framework that RM has.
There is an agreed upon method: use object identifiers. Object identifiers in OO play the exact same role that surrogates do in RM. It is a mistake to equate pointers with object identifiers because pointers are an implementation mechanism while object identifiers have a counterpart in logic. Surrogates--or any unary key for that matter-- have the exact same counterpart in logic: individual constants.
> > Since I clarified my position in a later post, I have to wonder what
> > you sought to gain by misrepresenting what I wrote. But then again, I
> > should give you the benefit of the doubt: maybe you didn't read the
> > clarification.
> That's correct, I don't lurk here all the time, and I probably missed
> some articles in the (long) thread. Perhaps my news server has even
> dropped some.
> Clifford Heath, Data Constellation,http://dataconstellation.com
> Agile Information Management and Design
Received on Mon Sep 28 2009 - 20:22:52 CEST