JW – I’ll probably be giving you a talk slot at the annual meeting. Expect 20 +- 10 minutes.
MT – Ok.
Trello
MT – Had some trouble with DEXP FF where it had an empty vdW section.
JW – I don’t think an empty vdW section is valid.
(some discussion)
The relevant section has a wildcard parameter that will always have 0 energy, not an empty section
Issues and PRs
What other changes can MT make in OFFTK
What changes in general can MT make in Interchange without getting in trouble?
JW – Basically I think the “simple” answers are:
“Do whatever you want in interchange but try not to break stuff”
“Change whatever you want but ask me if it’ll change evaluated energies”
MT – Neither is very good. I want to have a test suite or some sort of button to press that will tell me whether a change is acceptable. Right now I rerun the interchange regression tests.
JW – I think the interchange regression tests are sufficient for my purposes - I just want to avoid cases where OFFTK makes OMM systems with different energies without any notice, and I think these 10 mols are a sufficient test suite that we can add to if needed.
MT – And API breaks?
JW – I think interchange is all yours, I mostly just care that I get ~1 month of notice if the API will be breaking so I can get the toolkit updated. But if you get frustrated users from behavior/API changes, they’ll come directly to you.
MT – I want to make sure I avoid flak for something like AHogan’s stuff breaking, so I’m wondering what guidelines I can follow to avoid that. I also want to avoid people saying “oh, that sounds interesteing, I should try that” without actually giving me feedback.
JW – I’ll add this to the “hard decisions” for this year’s roadmap planning - Basically that we’d like to know our relationship to researchers and expectations, given that many of them are using old stuff/unversioned branches of things.