On Wed, Oct 04, 2006 at 11:45:49AM -0400, Stavros Macrakis wrote:
> Well, there are two issues here:
>
> 1) What definition to use for round? Round towards zero or towards minf?
> Towards nearest even (numerical analysts' definition) or simply
> truncate(x+1/2) (traditional defn)?
I suggest a switch that allows choosing the global behavior as either
round-to-even or truncate(x+1/2) with the default being the truncate
definition which more people will be familiar with.
--
Daniel Lakeland
dlakelan at street-artists.org
http://www.street-artists.org/~dlakelan