Evan Pretti

I believe that in the implementation of this kind of barostat in [LAMMPS](https://docs.lammps.org/fix_nh.html), one can specify any symmetric stress tensor, so it can have non-zero off-diagonal elements as well as...

It might also be useful to report which precision modes are supported by each device.

Could `getDevices()` take an optional precision mode argument that, if present, filters the list of available devices returned?

I'm surprised to see that error as I thought it should have been fixed for Espaloma by this commit: https://github.com/openmm/openmmforcefields/commit/722c2f61b61b35974e4ac5d363f0a006d31ad342. If you have a self-contained example that reproduces the problem,...

The GAFF XML files distributed with openmmforcefields shouldn't be read by `GAFFTemplateGenerator`, so changing the 1-4 scaling in those files shouldn't have an effect. When I try to reproduce this...

> I think you've already explained part of my misunderstanding in your reply (i.e.the GAFF xml file is not read by the `gaff = GAFFTemplateGenerator(molecules=molecule, forcefield="gaff-2.11")` call, which I find...

For reference, this is partially resolved by #386: for the 20 standard proteinogenic amino acids and their ordinary N- and C-terminal patched variants, there should be no patch conflicts or...

Wanted to provide a quick update here. I wanted to see if we could fix some of the long-standing issues with improper ordering in Amber, since I was previously able...

Double precision for the OpenCL platform won't work on most Macs to my knowledge as the hardware just doesn't support it. I know this is true for the M4 Max...

I agree that we have some changes worth releasing (I think #390, #414, and #419 in particular). I think a `0.15.1` is fine, since we are really just updating the...