- add explicit dependence on the python-glue-ligolw-tools package

- switch from glue-segments to ligo-segments package

- elements must be installed using python3, but rest of this package
  requires python2 so we can't switch configure.ac over to it yet.  put a
  temporary hack into the gst/python/Makefile.am script to make it use the
  python3 interpreter during module install and byte-compile

gstlal-calibration/tests/check_calibration:  Added GPS start time to name of files in pcal-to-darm plotting script.

gstlal_inspiral: set the horizon distance calculation to 15 Hz instead of 10Hz since 10hz doesn't work for SEOBNR_ROM in the bns mass region

gstlal_inspiral_marginalize_likelihoods_online: FIXME: for now lower the number of samples per bin to 40,000 giving about 10 million for the O2 bank.  We could use more but this program will need to be parallelized.

gstlal_ll_inspiral_state: FIXME this implements a hack to take into account that ALIGO uses a 16Hz dq vector and virgo uses a 1 Hz dq vector. gstlal inspiral should just report the output of added and dropped data in seconds rather than samples and this would go away

- the point here, folks, is that we're trying to *untangle* the handler
  code.  a lot of work went into separating the eye candy related stuff
  into its own box.  the process of surgically separating the two is not
  complete, and the kafka patches are re-introducing entanglement.

- switch ranking statistic to this unconditionally for the October 2018
  open public alert test run.  after the test run and before O3 we will
  introduce more user-friendly plumbing allowing this to work with other
  template banks.
- NOTE:  git-lfs is used for any file ending in .hdf5 in the
  population_models/ tree

streamthinca, far: REVERT ME: A temporary home for the network snr cut which undoes the LR version and puts it back into the ntuple compare func. Once we have ironed out the kinks in the LR version, this should be undone

streamthinca, far: REVERT ME: A temporary home for the network snr cut which undoes the LR version and puts it back into the ntuple compare func. Once we have ironed out the kinks in the LR version, this should be undone

gstlal_ll_feature_extractor_pipe: add kafka/zookeeper nodes to DAG, update idq optimized makefile to deal with different cluster sites better, export kafka opts