The state of PTmon:  I think that I have probably fixed the fast rate of
sending fairy triggers.  Lately I have been tracking the background on time
intervals that are too short.  Over the summer, I spent a lot of time
examining data, and changed the frequency bands I am looking at on all
channels.  That meant reducing the downsampling rates so that the sampling
Nyquist would be above the upper limit of the frequency band.  That in turn
meant that without compensation, the time intervals for the data being fed
to the background checker were too short, and so I was triggering on every
little data quality-type glitch.  (A fixed number of peak-to-peak values is
fed in at a time, and so with an increased sampling rate arriving, this
number represents a shorter time interval.)  I have also suppressed
reporting of glitches below 5 sigma.  Everything looked good when I ran the
code briefly from Tech, but of course, I haven't been able to check the 16K
channels at LLO.  I hoped to run it this morning at LHO, where there is a
much better chance of the interferometer locking, but with the local network
down ... Maybe we'll be up this afternoon.