On Wednesday 13 December 2006 22:55, Robert Klemme wrote:
> On 13.12.2006 20:33, Tom Werner wrote:
> > And upon reflection, this makes sense, as in any sane world !(f1 ==
> > f2) will equal (f1 != f2), and so redefining != would be redundant.
Well, I remember a thread on ruby-lang where somebody were trying to build 
CSPs (constraint systems) using overloaded operators. In CSPs, you encode 
== and != differently and since you can't overload both in Ruby, he could 
not do what he wanted.

He could have in C++ (and somebody at my lab did ;-))
-- 
Sylvain Joyeux