> Barton's proposal seems to require some loop-avoidance fiddling, I think. > RJF Yeah, maybe the reason the author of check-noninteger-facts didn't make look at only one ply was fear of infinite loops. Bart