Dear all, we noticed that the way we determined the error bars on the unfolded cumulative distribution is incorrect. The spectrum itself and the moments are not affected. We fixed this and updated BAD 894 (the supporting document) and BAD 976 (the CONF draft) accordingly in CVS. Both are tagged with with tag KT-111604. For BAD 976 this means that two plots (int_976.eps and int_976belle.eps) got updated by us. The difference to the former versions of the plots *is* obvious. What should be done now? I apologize for not recognizing this earlier. Kerstin