Hmm, thanks for the explanation. I still think one could argue, however, that there's nothing wrong with using such logic when the issues you describe are 'trivial' (as you put it). If all the operations are perfectly reversible the p implies q also necesitates q implies p , does it not (correct...