[daip] VLBAFIX looping

[Jim Ulvestad]

I've run VLBAFIX a couple times today.  When I run
it on a data set that doesn't need FXPOL, it
quite happily goes and runs UVCOP multiple times
to split up the different frequency bands.

HOWEVER, when the first freqid needs to have FXPOL
run, because I observed RCP and LCP, but did not
correlate the crossed hands, it stops after
running FXPOL on the first FREQID.  The other
two FREQIDs never get processed.  It looks like
there's a loop that's not getting invoked in 
this case.  I took a quick look in VLBAUTIL.001,
but couldn't see where this was happening.

Observation is BU022; I'll try to leave the tape
some place obvious in my office for Monday, in
case Amy wants to borrow it.

j