[Kstars-devel] branches/KDE/3.5/kdeedu/kstars/kstars

[Jason Harris]

Hi Heiko,

> Try something like round(), there should be a function like this somewhere
> in the math-library. If there is none: try int(d+0.5), in order to
> implement round.
>
No, we don't want a round() function here.  If the value is 13.6, we want the 
integer part to be 13, not 14.  Again, the problem is:

double x = 13.00000000000;
int i=int(x);  // i == 12, *not* 13!
int j=int( float(x) ); // j == 13

I don't know why it works.  I would think that the double-precision value 
would be less susceptible to roundoff error, but it appears that casting as 
float before taking the int solves the problem.

-- 
KStars: KDE Planetarium
http://edu.kde.org/kstars