Thanks , when I follow that, the header is
function FloatToFrac(F: Double; Precision: Double): TFraction;
and I have that working from the other thread, but it seems to have less control than this
procedure FloatToFraction(AValue, APrecision: Double;
AMaxNumerator, AMaxDenominator: Int64;
out ANumerator, ADenominator: Int64);
but I think you are saying that header-related code is now gone ?
Mostly, I want to operate set fraction seek-depth by precision, but in some cases an upper limit on Denominator is also useful to provide.