Femtosecond synchronization of clocks

The precise-timing community is currently working on the task of synchronizing moving clocks in the field to an accuracy of femtoseconds. We argue that its recent claims of partial success must be examined in the context of relativity, and show that one must establish the context very carefully when claiming any sort of synchronization at this level. In particular, irrespective of the sense of the synchronization, we show that realistic movements of the clocks will destroy the synchronization moments after it is established.