Matt Thompson

Just a quick thought before I dig into this later - topology files need unique molecules and can't store information about copies, but coordinate files need all copies and don't...

Okay, there's a lot going on here and I need to be extra careful to only fix what needs fixing. I've opened a new issue to track requirements for various...

It looks like it worked fine until you changed the hydrogens?

The template-matching code has some requirements about the molecule's atoms and bond graph being the same when you call `createSystem` as when the template generator is constructed. Then, you modified...

Actually, your error is different than it appeared at first glance. You're re-creating multiple objects from different sources and assigning them the same name which confused me. This didn't throw...

I've been burnt by this several times over the years; it can get a pain to implement everywhere but also causes problems when it's not or only partially-implemented

``` In [10]: from openff.toolkit.utils.toolkits import RDKitToolkitWrapper ...: from urllib.request import urlopen ...: url = "https://www.ebi.ac.uk/chebi/saveStructure.do?sdf=true&chebiId=52 ...: 729&imageId=0" ...: rd = RDKitToolkitWrapper() ...: mol = rd.from_file_obj(urlopen(url), "sdf")[0] ...: print(mol) #...

I'm pretty sure the problem is that `OpenEyeToolkitWrapper` [_only uses itself_](https://github.com/openforcefield/openff-toolkit/blob/0.16.1/openff/toolkit/utils/openeye_wrapper.py#L2453-L2459) to generate conformers, implying that it can't fall back to RDKit for conformer generation as you'd expect `Molecule.generate_conformers` to...

I dug into this a little more and found some holes in how we handle users with some (but neither all nor none) OpenEye licenses. `OpenEyeToolkitWrapper` check at initialization if...

The decision-making power ultimately lies with you and the OpenMM team, but I have to caution a little bit against this specific combination of functionality. The farther an implementation with...