von neumann machine (Re: is squeak really object oriented ?)

Jecel Assumpcao Jr jecel at merlintec.com
Fri May 23 00:45:39 UTC 2003


On Thursday 22 May 2003 19:48, jan ziak wrote:
> i have read the webpage at the supplied location. i draw the
> following from it:
> - the automaton is constructed in such a way that IT SUPPORTS the
> reproduction process. thus, as i said, the organism (in my case it
> was a glass) must be situated in an environment which provides for
> its reproduction and thus no object can replicate itself as such.

Any object takes up resources to exist. Space and mass in the real 
world, for example. Once you reproduce it, the result must take twice 
as many resources. Since the first object didn't exist without the 
environment/resources, I don't understand why it bothers you that the 
new object needs them as well.

> - an empty cell in von neumann's machine has also its logic, its
> behavior.

Just as empty space "has" the laws of physics in the real world.

> so the question now is: are the empty cells copying the
> machine or is the machine copying itself? i thin that the answer to
> this question is that the machine exploits its environment in such a
> way so that it seems to us that it replicates itself.

The machine *is* copying itself by any reasonable definition, even 
though it might have to extract resources from its environment to do 
so.

A cell is a self-replicating machine. But not if it is floating in 
space. It can't even function in that situation, so I don't see the 
point of excluding the environment. A strand of DNA isn't a 
self-replicating machine (not even when wrapped in some proteins: a 
virus), so objects must have a certain level of complexity before they 
can copy themselves.

-- Jecel



More information about the Squeak-dev mailing list