General updates | JW – Making progress on Vsite tests, but scope-crept. Should I update on progress? Useful to make a Interchange WBS? Shirts group (?) crystal code (?) review (?) MT I’d be happy to take on the work associated with this. Envisioning a call with THuefner where I either help him port the code into the cookiecutter, or do it myself. Recommend not touching the keyboard/typing yourself, rather walking him through it so he gets an understanding. Aim for getting just a few integrations running on a basic level - Codecov, GH Actions testing, readthedocs.
I’ll be able to help port tests, not sure if I can port tests into existence If you end up needing to make the first tests, it’s fine if thy’re ridicuously simple, or if simple tests aren’t possible, just setting up 2 files with some assert 1+1==2 tests would be fine as well
Which member of the Shirts lab was it? Not sure that it’s a good idea to merge into the toolkit. Basically all ported code should go into a separate repo. Their needs may drive changes to the Toolkit, Evaluator, or Interchange, but it shouldn’t be assumed that we approve copying code in verbatim.
I might have time in a month or two. Large uncertaintly about whether this is a 1 hour, 10 hour or 100 hour project. Big unknowns are Do they plan to substantially rearrange classes? How extensively do they want to incorporate OpenFF functionality (answer is probably “wherever it makes sense”, but then either someone with OpenFF expertise needs to read all their code, or someone who’s writing their code needs to acquire openff expertise) MT will make contact with MS/TH and ask about availability for the first week of april. If scoped work looks like substantially more than 10 hours, MT will tell JW and we’ll figure out how to budget for it.
non-periodic PME is a PITA, now throwing exceptions Hack on Interchange.from_openmm?
|