Re: [boost] Infinite precision integer draft

From: Maarten Kronenburg (M.Kronenburg_at_[hidden])
Date: 2006-05-29 17:21:56

The user should set the modulus of the
unsigned_integer once in the beginning.
But for a mathematical modulo class,
the user can make a template with an
unsigned_integer data member,
and a non-type template parameter that
holds the modulus.
Upon instantiation of that template,
the unsigned_integer static set_modulus
function must be called with that modulus.
Or simply a static initialize() method in
that template which does that,
or a static variable of an init class,
see The C++ Programming Language,
3rd ed., 10.4.9.
The point is that then a template type is
associated with the modulus.
Regards, Maarten.

"Gerhard Wesp" <gwesp_at_[hidden]> wrote in message
news:20060529153226.GG29026_at_google.com...
> On Mon, May 29, 2006 at 05:15:59PM +0200, Carlo Wood wrote:
> > One for the infinite group of integers, and a templated one
> > (so basically, infinite number of types) for the finite groups.
>
> This has limited usefulness.  The more common applications of finite
> groups (crypto...) require moduli
> - well beyond of what's possible as a numeric template parameter and
> - determined at runtime.
>
> > and 'unsigned_integer' for the finite group (with a static
> > method to set the modulo). The only thing I disagree with
>
> Aaah, so I got this wrong all the time.  I thought Maarten's
> unsigned_integer should be, well, an unsigned integer, without a
> modulus involved.  So it should represent Z_p?  Maarten, can you
> clarify?
>
> Regards
> -Gerhard
> -- 
> Gerhard Wesp
> ZRH office voice: [phone]
> ZRH office fax: [phone]
> For the rest I claim that raw pointers must be abolished.
> _______________________________________________
> Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost
>