Re: operators.hpp issue

From: helmut.zeisel_at_[hidden]
Date: 2001-08-30 01:50:53

--- In boost_at_y..., "David Abrahams" <david.abrahams_at_r...> wrote:
> 
> ----- Original Message -----
> From: <helmut.zeisel_at_a...>
> 
> 
> > Long time ago I asked whether  operator.hpp could
> > support the other version of non-commutative operators like
> >
> > friend T operator-(const U& x, const T& y)
> > {
> >   T x1(x);
> >   return x1 -= y;
> > }
> 
> I think it's a good idea, though I don't remember if I had any 
objections
> earlier ;-).

Just to remember:

Your point was that this is not meaningful in every case:

"interator - int" makes sense;
"int - interator" does not make sense.

In addition, you pointed out that no existing code
must be broken,
so I added only new structs
and did not modify existing structs.

> If the intention of your email is to ask again whether your changes 
can be
> incorporated, then I'll ask again for a proposed patch (including
> documentation). A code snippet like you have supplied below is 
pretty good,
> but I'm sorry to say that I don't have time to figure out where it 
should go
> in the header. A patch would be much better.
> 

On Monday, I sent an updated version of operators.hpp
and operators.htm
(based on the version in boost_1_24_0)
per mail to you.
Since I did not receive any reaction,
I assume it is lost.

Now I uploaded the same version:

http://groups.yahoo.com/group/boost/files/big_int/operators.zip

If there are any question, please let me know soon,
in September I will be on leave.

Helmut