- add snr_min cut to remove sub-threshold triggers in case itac starts
  producing them

- use new "event_collector" feature of snglcoinc to identify singles
  suitable for use in background models.  this should fix the problem of
  the sum-of-SNR^2 cut introducing a bias into the noise model PDFs.  in
  any case it decouples background collection from other aspects of
  candidate construction, allowing for the selection criterion to be
  adjusted more easily, with less thought required.

- remove some code that was only there to make error messages less cryptic.

gstlal-calibration/tests:  Added to scripts to test how much nearby calibration lines contaminate each other as a function of frequency separation.

gstlal-calibration:  use FFTW_ESTIMATE instead of FFTW_MEASURE so that output does not depend on start time.

gstlal_feature_synchronizer + gstlal_ll_feature_extractor_pipe: fix issue where sample rate of incoming features was not being taken into account, expose option to online DAG generation

gstlal-calibration/tests:  Added more filtering options to plot_transfer_function.py so that it can handle whitened channels.

- gstlal_compute_strain:  numerous small bug fixes to get kappa_U calculation to work with new channels.
- tests/check_calibration:  Added plotting for GDS / CALCS transfer function.  Still needs dewhitening.
- debian/gstlal-calibration.install:  edits to make gstlal-calibration build on Debian

Update .gitlab-ci.yml: trying to build documentation with an older stretch container. editing in the web interface, because, yolo.

- use TYPICAL_HORIZON_DISTANCE constant instead of 100. for the typical
  horizon distance

- avoid evaluating float("-inf") in performance sensitive paths

- missing Dataless and Frankenstein ranking stats

- currently nonsense, redundant since addition of sum-of-SNR^2 cut.  the
  latter will be fixed, cleaned up, and made to incorporate this feature