[Haskell-cafe] Re: [off-topic] OOPer?

[Derek Elkins]

On Mon, 2008-11-17 at 16:04 -0800, Jonathan Cast wrote:
> On Mon, 2008-11-17 at 21:49 -0200, Maurí­cio wrote:
> > >  (...)  I don't recall where I found the following example, but copied
> >  > it locally as compelling evidence  that the functional solution can be
> >  > much clearer and  shorter than the same solution  modeled with objects
> >  > and inheritance.
> > 
> > Greg,
> > 
> > I desagree with you.  Bjarne Stroustrup, the original creator of C++, is
> > a sensible person
> 
> I think his creation of C++ is evidence against that view...
> 
> >  and I share his peacefull opinion in this matter:
> > 
> > http://www.research.att.com/~bs/bs_faq.html#compare
> > 
> > Even with good  intentions, I've never seen such  kind of comparison not
> > to fall into religious fights.
> 
> How do you recommend selecting a language?  Stroustrup, in my
> experience, seems to think the answer is `any method but technical
> merits', which would make me suspicious of the technical merits of his
> solution even if I knew nothing else about them (and I know all too
> much...).  Ask yourself: what is he hiding?
> 
> >  > -- Arithmetic expression forms data Expr = Num Int | Add Expr Expr
> >  >
> >  > -- Evaluate expressions
> >  > eval :: Expr -> Int
> >  > (...)
> > 
> >  > public abstract class Expr {
> >  >    public abstract int eval ();
> >  >    public abstract void modn(int v);
> > 
> > Although I'm not good enough to judge anyone's Haskell code, the Haskell
> > version  seems nice.  I  don't  know how  someone  who understands  well
> > object-oriented code  would do  that. But I  did C++ until  around 1998,
> > when the first standard was set, and  I can tell you for sure that, even
> > at that  time, no one who  knows at least  the basics of C++  would ever
> > write that problem like this.
> 
> Of course not.  But their solution wouldn't be object-oriented, either;
> this is just an example where OO solutions don't make sense.  (And the
> example code is Java; I've been wracking my brain, but I can't come up
> with any other way to do it in that language).

Jonathan is spot on.  The OO solution is an elegant approach.  I can
definitely see a C++ programmer writing it.  If they didn't write it
that way, it would be because, as Jonathan says, they didn't use an OOP
approach.

Note that this particular example is usually used to show the duality
between FP and OOP.  To add a prettyPrint function to the FP version,
one need simply write another function while the OOP version will
require changes to all the classes.  To add a Mul constructor to the OOP
version, one need simply write another class while the FP version will
require changes to all the functions.  This example illustrates (and is
the namesake of) the expression problem.