RE: [boost] Formal Review for Interval Library - which significandconstants needed?

From: Paul A. Bristow (boost_at_[hidden])
Date: 2002-09-08 17:26:06

Can you re-explain please.

As envisaged at present, only the value of numeric_limits<>::digits
is used to chose the right interval constants.

If this value is wrong tough.

Or not defined then won't compile.

This should make it work correctly even if long double is 53, 56, 64, 106,
112, 113, or 128 (?)

Or an I missing something?

Paul

> -----Original Message-----
> From: boost-bounces_at_[hidden]
> [mailto:boost-bounces_at_[hidden]]On Behalf Of Gabriel Dos Reis
> Sent: Sunday, September 08, 2002 1:13 PM
> To: boost_at_[hidden]
> Subject: Re: [boost] Formal Review for Interval Library - which
> significandconstants needed?
>
>
> "Paul A. Bristow" <boost_at_[hidden]> writes:
>
> | Anyone who needs to add a new one will also need to have
> | the arbitrary high precision arithmetic package
>
> Not necessarily.
>
> [...]
>
> | Can I also ask here for views on which are needed apart from the
> | obvious IEEE 24, 53, and 64, and 106 and 112 and 128?
>
> As explained earlier, if want to support long double, then you'll have
> to leave IEEE-754 requirements anyway.  And if the problem can be
> solved for that datatype, then it can be solved for float and double
> without the 24 and 53 bits requirements.
>
> -- Gaby
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost
>