"jan ziak" <ziakjan at host.sk> keeps the thread going.

This is positively my last posting on this topic.

	was it worth to learn all those languages ?
	
Certainly it was.  If you know only one language, you don't know that
language.  My native speech is English, and it's the only language
I'm fluent in.  But I wouldn't be as conscious of how English works,
or enjoy it half so much, if I hadn't picked up smatterings of others.
I can't _use_ any language other than English in practice, but it makes
me aware that things don't _have_ to be the way they are in English.
(Concrete example:  Maori and Tok Pijin distinguish between first
person inclusive and first person exclusive.  "yumi" in Tok Pijin
means me and thee, "mitufala" means me and some other person not thee.
The lack of this distinction in English was troubling to me as a child,
it was a distinction I often wanted.  But I didn't then know that it
was *possible* to do this in a word.)  And of course you don't have to
be fluent in a language to borrow words from it; where would I be in
this country if I didn't know what words like "whanau" and "iwi" meant,
or if I needed to go urgently and couldn't tell which door to go through
when one is labelled "Tane" and the other "Wahine"?

In the same way, learning several programming languages enlarges the
space of the thinkable.

	> And I honestly cannot see ANY useful resemblance between
	> computational objects and real physical objects.
	
	i personally do not share your opinion that (i cite) "i cannot
	see ANY usefull resemblance between computation and real
	physical objects".

	why not?  my mind is just an interpreter of what i sense,

What has that got to do with it?

	so why not to persuade myself that the things on
	my screen resemble to real objects?

But most computational objects are not on your screen.

Persuade yourself of anything you want to.
You can persuade youtself you are a poached egg, for all I care.

The question is not whether you can persuade yourself of something,
but where there is a useful resemblance between computational objects
and physical objects in general.

	i can persuade myself, when i see fish on a tv screen, that
	there is a fish somewhere in some see.  you cannot?

There is a causal connection between the flickering lights on your
television screen and a real object.  If there were no such causal
connection, then your self-persuasion would not be useful, it would
in fact be a delusion.
	
One of my colleagues has a screen saver which presents sea creatures
moving around.  If you were to look at at, and persaude yourself that
you were seeing images of real fish, you would be deluded.

In fact the fish on the screen *don't* resemble real fish in their
movements.  I find the screen saver in question annoying for that
reason; there are things that LOOK like fish but they don't ACT like
the fish they are supposed to represent.

Of course, even in the case of TV images, it is important to be aware
of the distinction between the real fish which once existed and the
images you are seeing NOW.  "I see an image of a fish on the screen"
does not mean "that fish now exists", it means only "there once lived
a fish which I am seeing an image of".


But all of this is irrelevant to the central point, which is jan ziak's
original claim that ALL objects should act like physical objects.

There has never been any dispute that things which are intended to model
physical objects should resemble the objects they are intended to model.
The question is whether EVERY object should look like something.

	> Why?  If I write a letter to my mother on paper with ink, I will have
	> *no* success trying to push it through a wire.  If I send her e-mail,
	> she'll get it in minutes through that same wire which was totally
	> impermeable to paper.  Making e-messages more like physical objects
	> would make them *less* useful, not *more*.

See the point here?  It is *useful* for a computational object (here an
e-message) to have a property that no physical object can have.

	i think you will agree that a wooden door is not a computation
	object.

	if i write write a letter to somebody on paper with ink, i will
	have no success to push it through closed door...

This is not just irrelevant to the point, it's actually a repetition
of part of my argument.  Physical objects have limitations; if you don't
model those limitations, your representations will not be persuasive
(like those jerkey "fish"); computatial objects are MORE useful if they
DON'T have those limitations.

	... so what is your objection against believing in seeing true
	objects on a computer screen.  (of course, i do not say that all
	objects should be considered as real objects)

But you did.  You said that Smalltalk isn't object oriented because
programmers use text instead of interacting with quasi-physical objects.

If you now say that some objects *shouldn't* have visual images
that pretend to be real things, then we have agreement and the whole
thread is revealed as pointless from the beginning.

And here I end.