> ...so i got the same result what Maxima returns. Looks like > Maxima is correct unless also Matlab has the same bug. Or rather, both systems do the exponentiation element by element:-( Ole