> In any event, I'm about to check in a fix that works correctly with conv2, > harmonic, and smith-normal-form. Great. It works on the PolySpace test as well. Thanks.