On 09/03/2013 03:03 AM, Erik Schnetter wrote: > Please disregard the previous patch -- it accidentally used 32 bits for double. > > The correct patch is attached. > > -erik > Seems fine to me. Regards, -Michele