hi blackgang
as someone mentioned here gps is accurate to 5m. if it's +-5m for 1 reading, we could potentially have an error of 20m for 2 readings up and downhill.
if the above scenario is correct, then a ruler with its inherent errors might not be too far off.
plus the cost of acquistion of a ruler is much lower than that of a gps. and i think i have a ruler lying around somewhere