<!doctype html public "-//W3C//DTD W3 HTML//EN">
<html><head><style type="text/css"><!--
blockquote, dl, ul, ol, li { padding-top: 0 ; padding-bottom: 0 }
--></style><title>Re: About explicit metaclasses and old experiences
wit</title></head><body>
<div>I had been using prototypes for representing "a kind of"
classes in Spacetalk in 1988 - 1989. The system consisted from a
lattice of objects (partial ordering => multiple dynamic
inheritance), objects representing "reusable behaviour" were
called "notions" (just like in the real world). The higher
the object in the lattice, the more abstract had been the particular
notion. More concrete notions had "specialized" the
notion/notions from which they inherited. "Instances" had
been atomic objects - leafs (while any object was still able to act as
instance - every notion was also an object).<br>
</div>
<div>The aim of the system had been to be able to represent (to model,
simulate) real-world dynamic systems (rather than to be a programing
environment). It had included not only attributes and methods for any
object, but also laws (rules), constraints, actions and events, body
of the objects (spatial objects) ... In fact, it had been implemented
in Smalltalk/V and later in Smalltalk/V286 (after having learned the
limits of Simula and having had a bad experience with Lisp). Later in
1992 - 1993 my students have ported parts of it to ParcPlace
Objectworks\Smalltalk 4.0.<br>
</div>
<div>I am still convinced that so far this was the best "real
world model". Still, it was aimed for "knowledge
representation and AI" - I assume that Squeak has more ambitions
to become the "top programming environment" rather than to
become an AI environment.<br>
</div>
<div>I have now a plan to develop an Internet-based object-oriented
knowledge representation environment - a kind of "new Spacetalk
for Internet" - real objects on the Internet; ... and to
implement it in Squeak. Hence my question:<br>
</div>
<ul>
<li><b>- what are the ambitions of Squeak in long-term :</b></ul>
<blockquote>
<ul type="square">
<li>- to be become the best programming environment (eventually a one
suitable for implementing OO KRS),<br>
- or to become the best object environment (eventually sacrificing the
class mechanizm to favor of prototypes because of the distribution and
Internet).<br>
</ul>
</blockquote>
<div>I am asking just because this is not the first case when similar
thoughts are being discussed here (e.g. a discussion on real-world
units some time ago).<br>
</div>
<div>Thanks<br>
</div>
<div>Augustin<br>
<br>
</div>
<div>Steve Dekorte wrote:<br>
</div>
<blockquote type="cite" cite>Wouldn't using prototypes eliminate the
need for metaclasses<br>
and make intsantiation and initialization more understandable<br>
to children(and adults)?<br>
</blockquote>
<blockquote type="cite" cite>Steve<br>
</blockquote>
<div> </div>
</body>
</html>