Re: I use Eiffel! and other "strange tongues"!

[walth@netcom.com (Walt Howard)]

On Fri, 8 Aug 1997 04:41:57 GMT, "W. Wesley Groleau x4923"
<wwgrol@pseserv3.fw.hac.com> wrote:

>> 1.  Many developers have had difficulty in learning OOP, and Steve
>> McConnell (in his book "Rapid Development") even goes so far as to
>> say that OOP is too difficult for the average programmer to do.

	I think he's right.

>OOP is difficult when OOP is defined and practiced by extremists in
>a way that makes it difficult.  This is diametrically opposed to the
>truth that OOP is often (NOT always) a natural way of viewing things.

	I would agree with the first statement. In most shops I've
been around, it takes an average C programmer 2 to 5 years to
wake up to OOP. A lot of them just don't get the point. 

	I mean really, 20% of programmers I've know don't even
understand how to break a program into functions! How many 800
line main()s have you seen? I've seen plenty. There is no way
that a function impaired programmer is going to grasp OOP which
is the next level up.

	The main hurdle I've found is that they mostly don't
understand why encapsulation is a good thing.

	Until they do, there is no way they'll take the effort to
design classes. Learning a new paradigm is difficult enough. If
you don't want to learn it because you think it's silly, it's
downright impossible.

	That is why programmers should learn a strict OOP language,
so they can't cheat. It FORCES them to design classes. Something
like Java might be good for C programmers. But they should learn
C first because learning programming should give one a brief
exposure to the way our modern paradigms were developed and why.
So learning the "old way" just enough to understand why the new
way is better is a good thing, otherwise, you never understand
how bad it was and why you have to take all the effort to do OOP
breakdown before you start coding.

	Walt Howard