Page Comparison

General updates

• MT –

  • Note: I think I’m not up to full productivity since annual meeting. Not sure why. Working on it.

    • JW – No worries, variations are natural.

• JW –

  • Just made a Fragmenter release , will let to fix a bug that appeared in Bespokefit, waiting on the conda-forge bot to open the PR

    • MT – Anything I need to know?

    • JW – Nope, was kinda fun but fairly deep.

  • Any smoke from MDTraj release testing?

    • MT – Nothing related to OpenFF. Some smoke elsewhere but we’re fixing that up. A few months ago Jessica Nash reached out to offer to help with MDTraj. She contributed a medium PR. I botched the merge - that’s being fixed now. JN+LNaden reached out to me soft-pitching a kind of MolSSI-supported MDTraj hackathon in the fall.

    • JW – Seems like 3-ish options:

      • We send you to this as a “conference”

      • You go to this as a part of your “unlimited vacation”

      • OpenFF takes on a formal role in this (but this gets jurisdictionally complex)

    • MT – MDTraj and OpenFF don’t have a ton of overlap. And I don’t know whether this is an in-person workshop or virtual. I think I’ll post about this more generally in OMSF slakc.

    • JW – Agree. I’ll bring up the “conference” option with lead team and see if there are any objections. It sounds like you don’t want to take on major organizer responsibilities here which I agree with. It would be helpful to tease out from them whether they have in mind a “maintainers summit” vs. a “new feature hackathon”, and what scope for participants they’re envisioning.

    • MT – Ok, I’ll respond to MolSSI email CCing JW+LW+DM.

    • JW – Great, I’ll bring this up with lead team.

  • After some thought, I think I’m going to unlock vsites to solve Cresset’s needs with a env var like OPENFF_UNSAFE_VSITES. Kinda stealing a page from the INTERCHANGE_EXPERIMENTAL approach. Do you have any thoughts?

    • MT – Sounds great.

Trello

https://trello.com/b/dzvFZnv4/infrastructure

• MT – Pydantic 2 migration ran into some fundamental changes. I think things are nearly done, but I think there’s a small remaining problem somewhere and it’s extremely hard to figure out where. It’s complex because we have a lot of custom models, and a lot of nested models. A lot of moving pieces because everything’s in different repos, I’m looking at solutions to simplify testing updates.

  • JW – Could pair program on this during Friday check in? I should probably learn how our stack is using pydantic.

  • MT – Yes, let’s do that.