[Mplayer-cvslog] CVS: main/libmpcodecs ve_xvid.c,1.18,1.19

[D Richard Felker III]

On Thu, Apr 03, 2003 at 09:25:39AM +0200, Dominik Mierzejewski wrote:
> On Thursday, 03 April 2003, Rémi Guyomarch CVS wrote:
> > Update of /cvsroot/mplayer/main/libmpcodecs
> > In directory mail:/var/tmp.root/cvs-serv28820/libmpcodecs
> > 
> > Modified Files:
> > 	ve_xvid.c 
> > Log Message:
> > 
> > Actually do something useful with XVID_GLOBAL_EXTRASTATS / XVID_EXTRASTATS.
> > Now we can really compare XviD and libavcodec :)
> > 
> > 
> > 
> > Index: ve_xvid.c
> > ===================================================================
> > RCS file: /cvsroot/mplayer/main/libmpcodecs/ve_xvid.c,v
> > retrieving revision 1.18
> > retrieving revision 1.19
> > diff -u -r1.18 -r1.19
> > --- ve_xvid.c	2 Apr 2003 22:27:35 -0000	1.18
> > +++ ve_xvid.c	3 Apr 2003 06:53:01 -0000	1.19
> [...]
> > +static double
> > +sse_to_PSNR(double sse, double pixels)
> > +{
> > +    return sse == 0 ? INFINITY : 4.34294481903251827652 * (11.08252709031685229249 - log(sse/pixels));
> > +    // 4.34294481903251827652 = 10/log(10)
> > +    // 11.08252709031685229249 = log(255*255)
> > +}
> > +
> 
> You should never compare real type variables with 0. Best thing to do in

You should never call float variables "real" types!!! Sorry this is
just a pet peeve of mine...a finite set of diadic rationals is about
as far from reals as you can get...

Rich