NetNews Usenet Archive 1992 #30/NN_1992_30.iso/spool/comp/lang/cplus/18109

Text File | 1992-12-16 | 1.9 KB | 42 lines

Newsgroups: comp.lang.c++
Path: sparky!uunet!microsoft!wingnut!pauljo
From: [e-mail] (Paul Johns)
Subject: Re: IS C++ a language for the "average programmer"
Message-ID: <[e-mail]>
Date: 16 Dec 92 21:18:12 GMT
Organization: Microsoft Corp.
References: <[e-mail]> <[e-mail]> <rmartin.724431509@thor>
Lines: 31

In article <[e-mail]> [e-mail] wrote:

> Yes, but does OO _really_ allow you to tackle a more complex problem?
> Or does it just allow you a new "notation" which is conveniently adapted
> to the more complex type of problems?  I've never heard of a programming
> goal which absolutely required OO to achieve it -- however, there are
> a large number of math problems which algebra just cannot solve, no
> matter how you manipulate it (assuming, of course, that you don't start
> making simplifying assumptions/approximations).

Given that people have a finite amount of time in which to work and
limited intellectual resources (especially speed and limited "short-
term memory") with which to work, a new convenient notation is very
important indeed if it allows them to solve more complex problems by
expressing them in quicker, more psychologically-efficient ways.

In the same fashion, one could ask, since all (compiled)
languages are translated into machine language anyway, do ANY
high-level langauges (or, for that matter, assemblers) "_really_
allow you to tackle a more complex problem?"

I do, however, agree that if you could simplify the process of
solving problems by ignoring the psychological characteristics
of humans (very limited short-term memory, for one) and the
fact that each person has a finite working time, then you could
effectively argue that it doesn't matter what language you use.

But none of us would return to programming by flipping front
panel switches....