Analysts: please note that the wrong schedule was 
used for the first ~20 min or so of the observation. 
The polarizations were mapped differently in the two
versions, so cjobgen may do something wacky during 
this time. 
LF