> > Today's problem seems to be only a problem for today... Also the push > > mirrors have the bad data, so it seems master started the push too > > early. > > The same script which runs the archive triggers the pushes, so that can't > happen. But the trouble happened nevertheless :-( I don't know what caused it, I Just can look at the results. Roman