From: Christoph <chr_news@...>
Date: 2002-04-02T04:30:28+09:00
Subject: Re: Why isn't Math object-oriented?

<nobu.nokada@softhome.net> wrote in
....
> At Mon, 1 Apr 2002 22:49:48 +0900,
> Mike Campbell wrote:
> > I'm new here, but I think the original question was not "how do I do it", but
> > "why is it like this", which Matz answered.
>
> Yes.  I just proposed an alternative without changing Ruby.

A plausible alternative is to change Ruby and reconstitute the current behavior in
the Math module.

module Math
    def sin(x);  x.sin end
    ....
end

From a practical point of view adding default container classes Set, Multiset,
and (un)ordered n-Tuples (these are easy to implement in Ruby of course)
might do a lot for the mathematical attractiveness of Ruby (maybe this should
be rubical attractiveness of Mathematics;-).

>
> > Still, I tend to agree with Ron; it seems unclean having a procedural set of
> > Math operations when most things aren't.
>
> But mathematics is not object oriented.

But at least some of the specialized mathematical scripting language like Macaulay-2
are kind of (claim to be?) object oriented - i.e. the categorical claim (the non-math
kind) that mathematics is not OO seems a bit strong to me.  A visible  stumbling block
of reconciling mathematics with ``single receiver-type OO''  is the often mathematically
unnatural  (I know a lot of people would probably think that this was a tautology)
distinction of method receiver and method argument(s).

/Christoph