Newsgroups: comp.lang.smalltalk
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!ix.netcom.com!netcom.com!rfenney.slip.netcom.com!user
From: rfenney@netcom.com (Robert J. Fenney)
Subject: Re: What makes a good Smalltalker?
Message-ID: <rfenney-210395120859@rfenney.slip.netcom.com>
Followup-To: comp.lang.smalltalk
Sender: netnews@mork.netcom.com
Organization: FenTek
References: <9503161937.AA27265@fred.nfuel.com> <mda94hta.11.000C21A5@rby.hk-r.se>
Date: Tue, 21 Mar 1995 20:09:05 GMT
Lines: 53

In article <mda94hta.11.000C21A5@rby.hk-r.se>, mda94hta@rby.hk-r.se (HANS
TAP) wrote:

> [...cut...]
> > I have argued
> >with quite a few that the real bueaty of Smalltalk to me is not that it is
> >dynamically compiled, or that it is typeless, or the memory management, etc,
> >but rather that I can approach the problem I am trying to solve with the
> >computer from a level that is familiar to me, rather than having to cast the
> >problem into some computeristic representation of arrays and stuff. Those
> >that get this, seem to take the time, as Ken points out, to take a step back
> >and analyze the actual problem.
> 
> >Travis Griggs
> >Siemens Power Corporation
> 
> I can only agree. I also belive that they miss one of the big points in the 
> object-oriented paradigm. One big problem is/was that software engineers used 
> several paradigms through the software development. The real world problem 
> domain was converted into an analysis model using one paradigm, the conversion 
> from the analys model  to the design model was done with another paradigm and 
> finaly the implementation model was based on a third paradigm. The problem 
> with this is that a model expressed in one paradigm do not relate easily 
> to the other.
> 
> Now, with the object-oriented paradigm you are abel to use the 
> same paradigm through the whole process. And I think Smalltalk is a good 
> language when it comes to the implementation part. You have a quite close 
> relation with the design modell. My point is that you decrease the 
> semantic-gap between the real world problem and the final application when you 
> use the same paradigm all the way through the development. If you use 
> object-oriented languages but still is stuck with the old way of thinking 
> (using multiple paradigms) you'll  have more a error-prone application.
> 
> One should try to decrease the gap between the real world paradigm and the 
> computer science paradigm as far as possible (without loosing performance)  to 
> get a more easily and error-free development method.
> 
> Hans Tap
> Student at the University of Karlskorona/Ronneby

It is funny but I think you may be missing a very important part to
developing in ST and that is the environments. The language is a great base
for building object libraries and tools which are then used to build the
next layer of components which are used.... Smalltalk buy it's self give
use nothing more than C++ offers it is the mature development environments
that give use the real advantage and the abillity to look at a problem with
a grandularity that is appropriate to the solution.

Just my 2 cents!


Robert
