From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.5-pre1 (2020-06-20) on
	ip-172-31-74-118.ec2.internal
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=3.0 tests=BAYES_00 autolearn=ham
	autolearn_force=no version=3.4.5-pre1
Date: 29 Aug 91 16:54:34 GMT
From: emery@mitre-bedford.arpa  (David Emery)
Subject: Re: c++ vs ada results
Message-ID: <EMERY.91Aug29095434@Dr_No.mitre.org>
List-Id: <comp.lang.ada>

The Booch Components ARE NOT a complete system.  There has been some
debate over whether the C++ components achieve the same level of
functionality as the Ada components, i.e.  is there a C++ component
for every Ada component.  

I would be very surprised if the Booch Components did not shrink in
SLOC when coded in C++.  They were designed using that kind of
object-oriented paradigm.  But, real systems consist of a lot more
than little data structure components.  

Where I work, 100k SLOC is a small program, 1-2m SLOC is average, and
a really big program might have 5-10m SLOC.  Ada is making a
difference in programs of this size (e.g. AFATDS, CCPDS-R,
STANFINS-R).  I've not heard of similar-sized C++ programs.  What I
have heard is that large C++ programs, particularly those developed by
different groups of people, and integrated, tend to fall apart when
something goes wrong in the inheritance heirarchy.  In particular,
trying to find out "whose method is this?" in a very large, complex
C++ program is damned hard.

By the way, if Darryl Cornish thinks that software is "finally
discovering the architect/engineer/trades-person paradigm" just now,
then he has missed much of the discussions in software engineering
over the last 50 years, both theoretical work (e.g. Chief Programmer
Teams) and practical experience.  Read "The Mythical Man-Month", and
then come back and say that architect/engineer/trades-person is a new
idea for software...

				dave