Hi, I'm looking at the new `linprog` routine that has just landed in Scipy 0.15. By coincidence, it just come on time for a new problem I have to solve these days ! While studying the signature, I found that the default argument bounds=None is a bit confusing. Indeed, if I got it right, bounds=None actually implies non negative variables. Is that correct ? However, bounds=None has a clear semantic meaning of "unbounded variables". In order to make this default choice more obvious (note that I'm not arguing for changing this default), wouldn't it be clearer to write `bounds=(0, None)` in the signature ? (maybe this was discussed already ? I didn't check Github...) best, Pierre