[Mnj] Polarization bug

[egreisen]

The handling of the parallactic angle correction has been in some
error since July 20.  This rotates the RL and LR to correct for
parallactic angle (and also adds any rotation measure correction
from TECOR).  This rotation happened whether or not DOPOL was true.
In the past and, now in the future, this will only happen when DOPOL
is true.  In this interval, July 20 - September 15, it was possible
to have the parallactic correction applied more than once, resulting
in bad values for Q and U.  The corrected code is now available for
download by the MNJ.  Note - this error only applies to 31DEC22.

Sorry about this

Eric Greisen