2022-02-23 Horton/Wagner Check-in meeting notes

Followups from Simon

• The docs still lack a theoretical overview of what the framework is doing behind the scenes, so @Jeffrey Wagner you may want to ask the two Josh's to add this (either on the Fitting workflows page or as a new Overview page under the BESPOKE USER GUIDE section prior to release, or at least to work on it afterwards

  • @Joshua Horton will open an issue with a list of bullet points for the theoretical overview, will tag @Josh Mitchell (@yosh...)

  • @Josh Mitchell will flesh out theoretical guide (no more than half a page), open PR, request @Joshua Horton review

  • @Josh Mitchell will add releasehistory page to Bespokefit

• I'd also recommend getting #118 merged to avoid any potential confusion.

  • JH – Merged

• When cutting the release you may also want to make sure that a correct tag name is set as this was missed last time.

  • JH – I’ll try making the release this time.

  • JW – Feel free to reuse some of the instructions from here - Step 2 has stuff that I copy and paste onto the GH release page

• Other than that I'd recommend making the 0.1.0 release with Josh H. when he's back, trying to get someone new to follow the quick start guide with something like biphenyl as a test case (make sure to use --directory if using the run command when debugging so the outputs / logs aren't created in a tmp dir), and if that goes well, pass on the good news to Bill so he can start experimenting.

  • (JH + JW cut a new release – BespokeFit 0.1.0)

  • JW + MT will update the staged-recipes PR. Everything here is required, even the things called “optional”

Next steps

• JW – Chemper issue? Chemper can’t close rings

  • JH – This isn’t a showstopper - Even without closing rings the pattern is specific enough.

• JW – Note to self: When marking conda package pin to openff toolkit 0.10.X!!! This will keep biopolymer release from breaking everything immediately.

  • JH – Note that this may need a toolkit release with COOH fix. So I’ll need to cut 0.10.3 or do something else clever here. (Though COOH issue is fairly rare, so we may be able to get away without this)

•  

Other

• Dotson PRs on qcsubmit

  • JH – I’ll take a look at this