The failure
One system records that a sample was analysed at 14:04. Another records 13:51 for the same event, because their clocks differ. Reconstructing the sequence of events becomes an argument rather than a lookup, and the record's credibility suffers for a reason that has nothing to do with the science.
