Rahul Sundaram wrote: > I suspect this is largely because feedback on updates-testing has > been historically very poor. Poor feedback is still better than no feedback. (-: -- Rex