MGF Developer Forum: July 2023

[richardotis]

Agenda (Proposed)

Time (Proposed): July 14, 2023 - 9-10am PDT
Venue: Virtual (TBA)
Attendance: Open to Public

pycalphad

• Path to 1.0: https://github.com/pycalphad/pycalphad/projects/2
  -- How ready does Workspace need to be in order to ship? (continuation of in-person discussion between me and @bocklund last week)
  -- Prospects for advanced phase diagram mapping prototype; release as a separate package versus integration, also does it need to integrate with Workspace or is it okay if it doesn't (yet)?
• Paper using pycalphad published in npj Computational Materials: https://www.nature.com/articles/s41524-023-01058-9
  -- Associated code: https://zenodo.org/record/7767663
• Sam Krimmel presented on space charge models implemented with pycalphad; how much of this work can be merged into mainline?

ESPEI

• ?? [needs project input]

Kawin

• Kawin paper published: https://www.sciencedirect.com/science/article/abs/pii/S1359645423003191
• Regarding paraequilibrium calculations, there's an interesting new paper that may be relevant for future implementation: https://link.springer.com/article/10.1007/s11669-022-00969-2

Scheil

• Any lingering compatibility issues now that recent PRs have merged? Any architectural changes that should be made to ease this issue for the future?

PySIPFENN

• ?? [needs project input]

DFTTK

• Maintainership currently in transition for this project, Penn State is working on it, who is the interim POC?
• What can MGF do to help?

Multi-Project Issues

• Where should MGF-affiliated projects live on GitHub? Is there a need to insist on the MGF org, or is it okay if they are disparate?
• Communication between developers, between developers and users, and users with each other: Discord versus Matrix versus something else
• How do we decide on this meeting agenda in the future?

Open Discussion

• Time allowing

@tonyfast This is what I was thinking for the developer meeting. I welcome your input based on your experiences running these types of meetings.
@bocklund I'd appreciate your help wrangling the LLNL crowd for putting together the agenda, as well as gauging attendance. Tony's concept is to run this meeting for an hour, followed by an open hour of demos from anybody (doesn't have to be us) on anything software-related, where we'd let people sign up for 5 minute slots in advance. I was thinking to schedule on RDO Fridays for convenience, but if there will be substantial LLNL attendance we could try to find something during the work week.

[richardotis]

July 14 may be fine for this one, but in general this might overlap too much with the regularly scheduled MGF board meeting. This plus demo day plus the board meeting could end up being 3+ hours of straight meetings every other week.

[tonyfast]

yea, i think we'd want to restrict our ask of folks to 2 hours in a day. lets plan on our first demo day next friday, and y'all have your regularly scheduled meeting. we'll have a demo day and developer forum the following friday (7/21).

if we had a google workspace we could start a community calendar. this is jupyters solution https://discourse.jupyter.org/t/jupyter-community-calendar/2485

[tonyfast]

i opened up a pull request to add these agendas to the repo #6

usually the agenda gets added to a repo after an event. this way attendees and notes can be captured in the minutes. that said, it would be nice to have a collaborative document for the meeting that we can add later. i've always used hackmd. with hackmd, we tend to keep a running for a a few months at time. it is a good centralized place to track what has been going on.

i'd recommend we review these agendas, and if they look right, then we use a service like hackmd to distribute them. curious if you have thoughts on a more acceptable way to collaborate with the community.