Richard Smith wrote: > (All I have changed is "if all such groupings are true" to "if all such > groupings are true and complete", which is arguably tautologous, but probably > clearer than simply "complete".) Ignore that: that described an earlier, less liberal version of the change. RAS