MT updates | Toolkit 0.10.4 release Seek to have ELF10 bugfix in Stereochemistry patch in? Any other critical bugfixes to include? Assume answer is “no” if nothing immediately comes to mind I can do/lead this
Regression tests: WBO molecules passing with new code Working on reproducing a possible issue with idivf in proper torsions MT – Seeing the same result from the regression tests that SB saw, but unable so far to detect this from scratch. (Screenshared on this, evaluated current TK behavior and implemented handling for ImproperTorsions idivf=None cases)
Switching function bug fixed in Interchange Yet to run full tests for small molecule or condense phase systems Individual runs are not too slow, but need to be re-run whenever anything is updated
Happy with Mitchell’s contributions/PRs: |
| JW – Thanks for pushing mbatgh to move issue to OpenFF Toolkit JW – I’ll be asking JHorton for main control over bespokefit maint and user support. JW – Can we keep return_topology? MT – My big objection to this is that is messes up the function signature badly. It would be easy to support this, but longer term it would be healthier to tell the user to create an Interchange in this case. Here’s what that could look like if it must be shoehorned into here: JW – I agree that it’s super messy, and that everyone should go through interchange in the future. Since maintaining reverse compatibility is so straightforward in this case, let’s shoehorn in support for the 0.11 series, but have it raise a deprecation warning, and completely remove it in 0.12. MT – Agree, I’ll do that.
|