DEV SESSION [TUESDAY] 14 May -- 2030 UTC+8/1430 CEST/1330 BST/ 0830 ET

[tpike3]

Where: via Matrix

Proposed Agenda

• GSoC discussion
  • GSoC Mentor Summit: https://sites.google.com/view/2024mentorsummit/home
• PRs:
  • add group_by method to AgentSet for attribute-based grouping #2092
  • Make connections in experiment cell spaces named  #2096
  • Minor Typos #2100
  • Fix deterministic behavior in move_agent_to_one_of with selection="closest" #2118
• Mesa 3.0
• Mesa-Geo
• Paper?
• Epistemix - https://www.epistemix.com/

cc @tpike3, @EwoutH, @Corvince , @wang-boyu , @rht, @quaquel

[jackiekazil]

High level notes from this meeting:

Discussion: Mesa and Mesa-Geo tutorial

GSoC Discussion:

• Dong: Can I extend for two weeks - and we said would
• Harsh: need time, not compute
• GSoC Summit: Jackie will commit to going.

Paper

• Andrew - thinks it is a good time for another Mesa paper -- Since the last one was 2020.
• Papers
  • Mesa 3.0 paper -- Summarize updates. (Boyu)
  • Main Mesa citation -- JOSS paper (Jackie)
• GSOC Papers?

More discussion to happen around papers.

No tickets were discussed because of time constraints.

[EwoutH]

Thanks for the notes. I agree on the papers. I was also discussing this with some other maintainers a while ago. I would be interested in contributing on the papers, especially on the PRs I worked on (AgentSet, PropertyLayers, scheduling).

I would frame it as a “Mesa 3.0” paper.

[wang-boyu]

Thanks @EwoutH. I was hoping to talk to you in this dev session about the paper. Is there a link to your past discussions about it? Will you be there in the next dev session?

[jackiekazil]

@EwoutH There are two papers here. I updated the notes to clarify.
One is the Mesa 3.0 paper, which is a summary of updates. The other one is Joss paper, which simply becomes the main citation for Mesa, so we have a consistent citation that people point, which holds the basic concepts of Mesa and that can been updated if needed.

[EwoutH]

It was somewhere on matrix, probably in private, don't remember where exactly. But the main idea aligns with the Mesa 3.0 paper Jackie describes.

[jackiekazil]

Closing this for clean up.