Integers and other numbers [was Re: dlw_int review]

From: Eric Ford (eford_at_[hidden])
Date: 2001-09-27 11:17:20

>  >> The only substantive suggestion...  If I wanted to tell
integer.hpp
>  >> that I have one or more "big_int" types that can hold larger
numbers,
>  >> I beleive I'd have to replace code a fair bit of code.  It might
be
>  >> nice to allow integer.hpp to be used with additional types
provided by
>  >> the user.  I haven't worked out the details, but I wonder if a
>  >> linked-list or tuple template parameter (with default template
>  >> parameter provided to function as is) could allow for this.
> 
> That's correct.  Note the 1999 copyright date.  integer.hpp was done
early 
> in the Boost effort when we had a much smaller number of members.  I
don't 
> remember any "big_int" discussion.

If a revised implementation were written that preserved the old
interface, but had a new implementation which added functionality via
an additional optional template parameter, would that be something
boost would be wiilling to consider.

> It would be nice if someone (or several people) with a strong
numerics 
> background could write a "big picture" roadmap that fits a bunch of 
> numerical type together into a coherent plan.  The ones in the C++ 
> standard, the C standard, and Boost as it stands today, but also
with slots 
> open for big integers, fixed point types, and other possibilities.

I'm not really sure what you have in mind.  In any case, I don't think
I'm qualified to write down the big picture, but I'd be happy to look
over a draft and tell someone they left out a piece that's near and
dear to my domains.  

If you're thinking primarily of numeric types, there's lots more...
rationals, arbitrary precision, dimensioned quantities, quantities
with units, and scaled quantities.  Not to mention types that involve
multiple elements...  

E