Re: wrong rhythmic dots

[Werner LEMBERG]

> > % . The algorithm for computing position of rhythmic dots
> > %   sometimes produces wrong results.
> 
> I can't duplicate this. Can you investigate?

I suspect a rounding problem somewhere -- perhaps a similar fix to the
one you've applied a few weeks ago would do the trick.  As you know, I
still use gcc 2.8 with an old libc version... a good testbed to find
peculiarities and border situations :-)


    Werner