General updates | MT – JW – Reschedule the weeks of sep 11 and 18 Should we take an alternate approach to FB updating? I’d held a lot of mental state internally assuming I was working solo but could we could sync up if you want to be involved. MT – Open to this but not sure it’ll help. History has shown that this is a massive effort sink. It’s a different software dev philosophy and I have trouble operating there. I think you put too much emphasis on letting LPW do things his way that the codebase is unacceptably far from best practice. Not sure there’s a path forward, and I’m not sure how to facilitate reducing tension/friction. Would propose removing myself from the situation. If I were dictator of everything I’d remove gromacs from the CI. The gromacs release may be months out. JW – Agree w/ you stepping back from it. Plan is to get tests green at all (hopefully just requiring gromacs), run nightlies, remove OE dependence, support new python versions. Agree this may take a very long time.
If you’re willing, I’d to to be updated on job search/ETD. WIll help a lot with cross-training timelines and balancing finances. pAPRika maintainers want to make a new release. I’m interested to train Jake Anderson (gilson lab researcher for a year) but am worried that a new release might mess stuff up for us. Do you see ways where this would make our life hard? MT – I assume paprika depends on openff. Evaluator has unit tests that interface with paprtika. I don’t think paprika depends on yank (though JSetiadi complained when I removed YANK interface)
MT – I’m willing to offer an hour or two of my time to work with JAnderson on this, would want a tour of changed code paths. MT – Last time I discussed with LW, pAPRika isn’t something that is important to OpenFF’s fitting or benchmarking. One thought is to help with a new pAPRika release and also remove the pAPRika interface from evaluator. JW – That’s a good idea. Would clean our hands of any scientific issues that come up from the changes (either to the interface or to pAPRika core). But would be “politically” unpopular MT – There’s a kernel of something important about our open source working style here - there’s one approach to open source that needs high reliability and active maintenance, and another that is open to all comers but has some thing collecting dust - and we’re identifying here that we’re in the first camp, and things like this are when we have friction with people who see our org as being in the second. JW – I’ll bring this up to lead team and will let you know decision. MT – I’m happy to spend 60-90 mins on helping out/mentoring regardless.
|