Merge branch 'bjmorgan:master' into COG

docs/source/faq.md

@@ -44,7 +44,5 @@
   > When calculating displacements, `kinisi` uses a simple heuristic to unwrap trajectories. 
   > If the displacement between two steps, is greater than half the simulation cell length, `kinisi` wraps that
   > displacement. This scheme assumes that no particle moves more than one cell between steps. Therefore, it requires that
-  > enough simulation data is provided to `kinisi`. This process is performed for each dimension of the trajectory,
-  > allowing for any orthorhombic (and cubic) cell. However, this heuristic does not support simulation cells that do not have
-  > lattice angles not equal to 90°, or cells that change shape or size.
+  > enough simulation data is provided to `kinisi`. 
   > This is the reason for not supporting NPT simulations, although this is being investigated.