Building Open Source Software for Climate Change Research — Lessons Learned fro...
summary
Mimi.jl is an embedded DSL in Julia that modularizes integrated assessment models and is actively used to produce social cost estimates used in EPA analyses and academic papers. The project relies on macros and Julia multiple dispatch and Julia package management for reproducibility; it has been adopted by several universities and government teams. A contextual inquiry study of domain experts finds the blurring of host language and DSL causes confusing debugging, so the team runs EDSL-specific forums and tutorials and promotes code reuse.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:27Talk overview and Mimi project
- 3:12Integrated assessment models
- 8:05Climate model structure
- 10:57Embedded DSLs in research
- 13:47Introducing Mimi
- 18:41Mimi's design choices
- 20:57Usability study motivation
- 22:23Usability study methodology
- 25:29Host language impact findings
- 26:56Debugging blurred boundaries
- 31:52Community engagement
- 33:19Designing for communities
- 35:06Code and template reliance
- 36:51Final design and conclusions
- 37:56Summary and Q&A
related talks
28mEnergy Models That Speak the Same Language: The Story of ESDL
29mPowSyBl: TSO Cooperation Through Building Blocks
40mInstructLab: Applying Open Source Methods to Building and Trai...
28mA community based open source LLM and what it can do for your business
49mDoing for Sustainability, What Open Source Did for Software
42mOS-Climate: A Breakthrough Open Data and Open Source Analytics Collaboration for...
42mOpen Source LLMs — Powering AI Accessibility and Innovation- 40mInstructLab: Applying Open Source Methods to Building and Trai...