GitLab

Using a const Vec3& instead of three doubles for the AnisotropicBarostat API.  Still figuring out SWIG issues.

Bug fix to Modeller

Improved error checking

Implemented RPMD ring polymer contraction

Add support for long long shuffle

Improved infrastructure for integrators

Integrators control when atom reordering gets done, instead of having it happen automatically when forces are evaluated.  This simplifies some code, and also improves unintuitive behavior with CustomIntegrator.

Bug fixes to use of shuffle

Completed implementation of DrudeSCFIntegrator.  Also cleaned up other details of Drude plugin.

Nonbonded.cu now uses shuffle intrinsic