From: Richard Karpinski
Subject: Re: Object IDs are good ( was: Object IDs are bad )
Date:
Message-ID: <dickkE9B98y.61C@netcom.com>
[newsgroups trimmed] Matthias Blume writes:
>There are things in the "real" world that don't have an "identity".
>Numbers are one example, and we generally don't expect to be able to
>distinguish between _this_ 1 and _that_ 1.
I agree so far, but then he writes:
>Another example is the electron (or other thingies that
>particle physicists might be interested in).
Sometimes --in some programs-- we care only about the number of
electrons in some region of space, and then we don't need to
distinguish between the electrons. But in other programs we
have to keep track of the position of individual electrons, and
when we do that we are distinguishing between them, IMO.
A deeper question: what do we mean by ``Object'' in the phrase
``Object IDs are good'' (or ``bad'')? Is an object a region
of the computer's memory? If so, then why have we said
``object ID'' instead of the more precise term ``pointer''?
From: Thant Tessman
Subject: Re: Object IDs are good ( was: Object IDs are bad )
Date:
Message-ID: <3363E1E5.41C6@nospam.acm.org>
Richard Karpinski wrote:
[...electrons...]
> A deeper question: what do we mean by ``Object'' in the phrase
> ``Object IDs are good'' (or ``bad'')? Is an object a region
> of the computer's memory? If so, then why have we said
> ``object ID'' instead of the more precise term ``pointer''?
The definition I like is that an object packages state. An
electron's position is a value, not an object. Using a
function to tranform the electron's position produces a new
value.
On the other hand, an object packages values in such a way
that if two parts of the program are looking at the same object,
one part can change the values contained in the object without
the other part knowing it.
Whether this is a good or bad thing is (obviously) the subject
of much debate. I've found that when I write in SML, most
(but not all) of the code I write winds up being functional
without me consiously trying to make it functional.
-thant
> From: ·····@netcom.com (Richard Karpinski)
> [newsgroups trimmed] Matthias Blume writes:
>>There are things in the "real" world that don't have an "identity".
>>Numbers are one example, and we generally don't expect to be able to
>>distinguish between _this_ 1 and _that_ 1.
> I agree so far, but then he writes:
>>Another example is the electron (or other thingies that
>>particle physicists might be interested in).
> Sometimes --in some programs-- we care only about the number of
> electrons in some region of space, and then we don't need to
> distinguish between the electrons. But in other programs we
> have to keep track of the position of individual electrons, and
> when we do that we are distinguishing between them, IMO.
This is not possible. Electrons have to be described by probability
distributions. If you have 2 electrons, the distributions overlapp
and the distinction (if there ever was any) between the two electrons
is lost.
If your program keeps track of electron-positions it is wrong...
There is a fantasy by the physicist John Wheeler: Why are all electrons
in the universe indistinguishable.
Answer: There is really only one electron, which moves like mad
through space-time and gives the impression of many electrons ;-)
In article <····································@ftn.allgaeu.org>,
···············@f16.n300.z53.ftn.allgaeu.org (Dieter Menszner) wrote:
> There is a fantasy by the physicist John Wheeler: Why are all electrons
> in the universe indistinguishable.
> Answer: There is really only one electron, which moves like mad
> through space-time and gives the impression of many electrons ;-)
Actually, it was Dirac who first suggested this. I attended a lecture
by Dirac when I was in high school and this idea was the only thing that
I remember comprehending at the time.