Re: Qualified sub-procedures

[Hans Boldt <boldt@xxxxxxxxxx>]

M. Lazarus wrote:
>  IBM Midrange history has taught us that it would be *very*
useful.  Years
> ago, what was the standard answer to the request to enhance RPG's
string
> handling?  Oh, if you want to do a lot of that use Cobol."  But I
know of
> no one that did that!  We have a large base of RPG programmers
that could
> ease into OO instead of going full tilt into it.  There is a
large learning
> curve that would be eliminated.  On top of that, RPG's I/O
processing (and
> many other features) beats Java's any day.
>

Booth Martin wrote:
> I am curious about something.  OO is Nirvana of course, but why?  What are
> the benefit(s) about OO that make it so superb?
>
> I ask the question because I wonder if we can do mock-OO with RPG's current
> capabilities, and gain the benefits?   We don't truly care about the
> features of OO, only the benefits, right?
I'll reply to both Mark and Booth since the questions seem related.

What is OO? After studying the subject for the past, oh, 15 years,
I'm not really sure I can give an answer. You can define OO in terms
of big $5 words like encapsulation and inheritance and polymorphism,
but a dictionary definition really doesn't say what it *is* and what
it *means*.

Mark, I suppose what I'm saying is that I don't think it's something
a procedural programmer can *ease* into. You can study it all you
want, but like Zen, sometimes you need an "Aha!" experience to
understand it, or at least some aspect of it.

I think the best way to understand OO and it's benefits is to study
the issue of "Design Patterns". Here's a book that offers a good
introduction to the subject. This explains the benefits of OO a lot
better than I can.

Design Patterns Explained
By Allan Shalloway, James R. Trott
July 2001
0201715945, paperback, 368 pages
Addison-Wesley Pub Co;

I think for a lot of people, OO is largely just a "buzz word". We
know it's good for us and can provide benefits. But often the
benefits are hard to quantify. In practice, sensible design with
flexibility and reuse in mind, even in procedural programming, is
really what matters.

In summary, personally, I think OO is good and worth studying. But
it's not a panacea for all our programming problems.

Cheers!  Hans