Switched to a slightly faster (and more accurate) approximation for erfc()
Fixed a compilation error in double precision mode
Prevent nan in CPU platform when two excluded particles are exactly on top of each other
Rewrote the neighbor list generation kernel to be simpler and (slightly) faster
Fixed bug in CustomManyParticleForce that caused crash
Bug fix to addExtraParticles()
Added a missing return statement
Created CustomManyParticleForce
Added scripts for generating force field XML files
Clean up some methods that serve no purpose in OpenMM.
Improve warning messages coming from the CHARMM parsers.
small typo in checkpointreporter.py true --> True