MT – Merged solvation example into interchange. Also set combine_nonbonded_forces to True. Will plan to make a release early next week, or after thanksgiving if I don’t get to it. JW – LPW meeting prep - Can we consolidate things into discrete changes? MT – Question may be “how do we keep progress from being limited by LPW’s attention?” JW – Would be good to define scope+liability for changes that JW approves+releases MT – I’d like to understand LPW’s future development plans, and how they align with ours. I think we’d struggle to compare deliverables+roadmaps because he probably doesn’t really have those. JW – Would be nice if there was a plugin-like solution here, where we could just maintain our side of it at whatever velocity we want. MT – I think the core issue that that FB is just crufty and unmaintained. Even if we maintain our own side of the plugin, problems arising from FB’s lack of maintenance will become our problems. JW – I’ll ask LPW about what he’d be willing to reorganize/trim off, and we can see how well that aligns with what we’d remove.
JW – Combine_nonbonded_forces - relationship to OMM 8 rc issue? MT – This is something that I think should be done regardless. I’m really happy with how the interchange development went, with nonbonded forces being separate, since this is the architecture we’ll ultimately need for alternative nonbonded implementation. But the early change was just kinda related to how we’re the only people on earth using switching in a customnonbondedforce. (JW realizes that the toolkit showcase explicitly calls Interchange.from_openmm and not forcefield.create_openmm_system, so this change WILL have a good chance of fixing it)
JW – OrderedDict PR
|