General updates | JW – Expect OFFTK 0.14 this week, mostly API deprecations. My follow ups include: inform fitting team about OFFTK version constraint for working with FB communicate to LPW the need to update FB pin bespokefit to old OFFTK (it looks like openff-forcebalance is getting different numerical results, so we can’t fully switch yet) MT – Agree, let’s move ahead with this. We can’t be responsible for maintaining our downstreams.
JW – JMitchell will merge centralized examples soon. He’s opened PRs on our repos to show how to get them into shape. JW – Will begin work on QCSubmit shortly after this. JW – New science team hires (Alexandra and Brent) start Monday - We should smother them with attention. I’ll try to find a way to get them to spend an hour or two with you next week (on conda setup screenshares or something). Please grab any such opportunities if you can. MT – Sounds good - This would probably be most natural coming from your or Lily initially. What are their roles/reporting structures? JW – Both report to LW. I think Alexandra is sponsored by the shared grant with MolSSI and will focus on QM datasets and valence fitting. I think Brent is sponsored by the NIH grant (but using OMSF as a subcontractor) and will focus on nonbonded fitting. MT – I may have some overlap with Brent if he’s looking at evaluator stuff.
JW – What are plans for packmol wrapper becoming public? MT – I think it’d be good to spin it into its own microrepo. Right now it’s just a small set of functions. JM suggested some good changes too. I’ll probably polish it for a few releases and if it stays well-behaved I’ll make it public in some form. Right now the big value to keeping it private is that we don’t have to keep API stable. What do you think of where it should live? JW – Philsophically it would fit better in the toolkit. On a technical cost/benefit level it would make sense in a microrepo (since then we could drop it without much fuss). MT – Largely agree, microrepo would make sense. JM thinks it should be in toolkit but system prep stuff is still an ongoing discussion. JW – Yeah, system prep is a slippery slope. MT – I’ll take the lead on future plans for this.
|