Newsgroups: comp.ai.shells
Subject: Why use a shell?
From: randy.bennett@dscmail.com (Randy Bennett)
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!europa.eng.gtefsd.com!newsxfer.itd.umich.edu!news.cic.net!ddsw1!godot.cc.duq.edu!newsfeed.pitt.edu!dsinc!netnews.upenn.edu!nova.voicenet.com!dscmail!randy.bennett
Distribution: world
Message-ID: <1.674.2209.0N27DE9C@dscmail.com>
References: <34qmot$86v@unix.sri.com>
Date: Tue, 13 Sep 94 23:03:00 -0640
Organization: DSC/Voicenet * Ivyland, PA * (215) 443-9434
Lines: 28

WO> Why use a shell to develop your KBS application?

Bob

        Everything I've done has been done with a shell.  My background
        is in industry, so the business case to justify the use of AI
        and the cost-benefit analysis are pretty important.  My company
        has done quite a few expert systems (even gotten some endless PR
        for some!) and all used a shell to one degree or another.

        We use a shell because we don't want to reinvent the wheel every
        time we start out.  Because we want to know that the reasoning
        and inference will happen the same way from application to
        application.

        Perhaps most important, we use a shell because we are business
        analysts who use computers and code to do our work, not coders.
        The use of a shell lets us focus on the business problem and
        less on the code.  We believe that the work of creating a
        workable, reliable shell, bet it expert system, neural net or
        whatever, has already been done.  We just take advantage of
        that.

Randy

---
  SLMR 2.1a  If it looks too easy, you must have missed something.

