GitLab

In calcForces() kReduceBornSumAndForces kernel is not needed since Born radii/Obc chain term now calculated prior to calculation of forces
Modified kCalculateObcGbsaBornSum kernel loop to match that in kCalculateCDLJObcGbsaForces1; Born sums are zero w/o mods

Decreased step size and simulation time

Optimization to constraint algorithm: skip the last iteration as soon as we see that all constraints have converged.

No commit message

Thanks to Chris, we now know that the Langevin friction parameter is in ONE OVER picoseconds, NOT picoseconds!!

Added doxygenable comment that Langevin friction coefficient is in units of collisions per picosecond (as opposed to picoseconds), which Sherm and I deduced by reading reference implementation and testing extreme values.

Checkin CMake for building the C wrappers. Checkin first attempt at Fortran 95 interface in the form of an OpenMM module. I don't know any way to build this with CMake.

Check in early attempt at NaCl example in Fortran; needs to be modified to match the latest C++ and C examples.

This is the "final" version of the HelloSodiumChloride example in C++. It has now been reorganized to put all the OpenMM calls in a few high-level functions whose interfaces do not use OpenMM. Also, the periodic box is gone with GBSA implicit solvation in place instead.

Per workshop dryrun, leave OpenMM:: namespace prefix explicit to clarify what's part of OpenMM and what's not.