Re: creaping coupling......

From: Mark Nicholls (nicholls.mark_at_mtvne.com)
Date: 02/24/05 

Date: Thu, 24 Feb 2005 14:43:36 -0000

"Thomas Gagne" <tgagne@wide-open-west.com> wrote in message
news:FZSdnf3r3cp6eIDfRVn-pQ@wideopenwest.com...
> It's problem like your's that give object oriented programming a bad
> reputation.

if its a genuine problem....and endemic (I don't think it's inherent), then
OOP tools deserve a bad rap, even if OO itself doesn't.

> Better to post struggles like this in comp.lang.java.programmer

maybe......its not a java question though....it applies to C#....and
probably java....and probably C++, and probably, VB6, delphi.....pick any
strongly typed lanuage.....it probably applies.....

> than comp.object. The OO naysayers don't hang out there and problems like
> this only give them more fodder.
>
> The only naysayers in c.l.j.p and c.l.j.a are those protesting against
static
> languages (which would also make your life easier).

The problem is fundamental in most strongly typed OO languages.....and here
in this group.....people (including those that design OOP's) continuously
intertwine class and interface, instantiation and operation and create an OO
mess and potentially give OO a bad rap........it is a serious OO
implementation problem.....unless I'm being stupid and there is a simple way
out.

What is the point of partitioning a system, if the tools we use force each
partitioned subsystem to know about the internal interfaces of all
others....what it doesn't have to.

It is not inherent in strongly typed languages, only in those (which may be
all), that for the owner of a typed object to know the intricacies of it's
interface methods.....