On Sat, Nov 25, 2006 at 01:45:43PM +0100, Michael Gerz wrote: > Georg Baum wrote: > >No, it does not fix a conversion, it suppresses a compiler warning. I > >don't understand why x_ is of type double at all. Why not int? > > I guess there are cases (like with multiple hfills), where x_ would be > incremented by a fraction of a pixel. If this happens several times and > we always strip the fraction, we get slightly incorrect output.
Right. Andre'