Should we drop support for loading in multiple XMLs?

[mattwthompson]

trafficstars

A summary of the question or discussion topic.

Right now you can instantiate a force field object with multiple files (Forcefield(forcefield_files=['ff1.xml', 'ff2.xml'])) which, as best I can understand it, is a holdover from how OpenMM assumes you have different XMLs for protein and water.

• I don't think we have a smart way of handling precedence (i.e. if water atom types are defined in each XML file, which should be used?)

• I don't think anybody actually uses this feature?

Therefore I think we should consider dropping it. Most of the code would remain the same, it would just be loading, processing, etc. one file instead of a list-like.