> > OK. > > (By the way, by "bad code" here, I meant code that performs badly, not > incorrect code.) I did write a couple of small test programs for run-time issues and did not see a problem. But if it makes sense, I can just use pointers to remove any uncertainties.