Rainer Deyke wrote: > Actually I suspect that integer division with negative values is rare > enough that a problem might escape notice for some time. Not for 25 years and a few hundred thousand users.