12 Feb
2009
12 Feb
'09
8:06 p.m.
on Thu Feb 12 2009, "Paul A. Bristow" <pbristow-AT-hetp.u-net.com> wrote: <snip massive quotation>
that it's too concise (i.e. cryptic), but I would say that these conversions and their C equivalents are sufficiently commonly used that users will quickly learn what they do. But apart from these syntax differences, are there any benefits to one version rather than the other?
What about handling user-defined types like big integer and big floating point like NTL ZZ, quad_float, RR, and GMP equivalents? Hardware is unlikely to go beyond 128 bit FP or int, if that, so we need to think ahead about more accurate integer and floating-point types in software.
Thanks, -- Dave Abrahams BoostPro Computing http://www.boostpro.com