Message 407024 - Python tracker
Mark, would it preferable to use ldexp() to build the float? + return math.ldexp(isqrt_frac_rto(n << -2 * q, m), q) - return isqrt_frac_rto(n << -2 * q, m) / (1 << -q)
Mark, would it preferable to use ldexp() to build the float? + return math.ldexp(isqrt_frac_rto(n << -2 * q, m), q) - return isqrt_frac_rto(n << -2 * q, m) / (1 << -q)