Jim Cochrane said:
[snip]
The question cannot be answered without a precise specification of the
OP's
requirements for the application to which the question applies. (Of
course, if there is no application, the question cannot be answered.)
Why not? Can't one define the terms 'state' and 'object' without reference
to a particular application? ie: I can define the term 'empty String' as
any Java String object x where x.equals("") returns true, without saying I'm
writing an XML parser for example.
And if you define these terms, how do you know the definitions will be
acceptable in terms of what the OP wants to do? I think that without
a context for the questions - requirements -, you're forced to make
assumptions - leading to endless philosphical and/or mathematical
speculations.
However, if the OP simply wants to know the answer to his questions outside
of a particular context, the answer is that there is no answer without a
context.
What if he just wants an answer to the specific question he asked?

To
answer the OP's original question, nothing but definitions of his terms are
required. ie: If he asked (for example) 'does String x = ""; do the same
thing as String x = null; ?' or 'what does the goto keyword do in Java?',
then no mention of a particular app or context (besides the fact the code is
in Java) is needed to answer these specific questions. On the other hand if
he asked, "What is the best way to design a GUI?", then the particular
application would be needed to answer properly.
Here's the original question:
"Does an object, which does not exist, have any state, e.g. the state
of not existing?"
My point is that there is no definite answer to his question, since he does
not supply any context or definitions, which are needed to provide an
answer. It could be answered conditionally - for example:
If we define existence of an object as having a state marked as
"existing", and non-existence of an object as having a state marked as
"not existing", then the answer is, yes, with these rules an object
has state.
But this is not very helpful, since he could easily figure this out for
himself. One could come up with other reasonable definitions for which the
answer is no; but then which one would be valid for what the OP has in
mind? He would have to decide for himself which set of definitions are
valid for what he has in mind - essentially, he has to answer the question
by himself.
If, on the other hand, he really meant: Is it reasonable to develop a
system for which an object that does not exist has state (the state of
not existing)?
Then I think the answer would be something like: Such a model sounds
reasonable and if it facilitated the development of a system in a timely
manner that meets its requirements, the answer is probably yes.
Anyway, it's probably time to stop bugging people with this OT subthread.
You can email me if you really need to respond. (See note below.)