If y and slope are floating point in the first block then might not y end at other than the desired endpoint because of accumulative error?