Continuous location updates

[jonls]

• Fixes Continuously update position from location provider #13 (also fixes OS X: there is no CFRunLoop #387).
• For Fallback location provider #36/Suggestion: configurable fallbacks #251/Add fallback location to redshift.conf #348/cache location information so redshift-gtk can be comfortably used with unstable internet connections #393/redshift-gtk: add an option to set default location (lat, lng) to use if no internet connection is available #396, we still don't cache the location on disk but if the location provider becomes unavailable, the last returned location is used until the location provider is able to provide updates again. In other words, at the moment we cache the last location in the running instance of redshift but not if redshift is restarted.
• This probably also fixes issues such as Startup crash if no working DNS #150, Could not get location / Provider does not have a valid location available #176, Redshift exits if no provider is found (no internet connection) #226, Unable to connect to GeoClue. Unable to get location from provider. #318, Redshift fails on startup #333 (though it doesn't fix the issues where Redshift cannot connect to GeoClue2 because the location services are not enabled/not allowed for Redshift).

[xeruf]

"we still don't cache the location on disk but if the location provider becomes unavailable, the last returned location is used until the location provider is able to provide updates again" WHY?

I just started my PC and had no internet, so I had to stare at this glaring white-blue because it couldn't track my location? Why not create a file with a few bytes in the tmp dir saving the last location?