Hello friends. I've been to a few of the London meetups over the years, but only just realised there's a whole online community! My name is Rory, I'm a PhD student in theoretical neuroscience at Cambridge, asking how biological neural networks use spikes and conductances to compute things. I used to be a software dev, and still consider myself a craftsperson first, scientist second. Right now my big interest is building better devtools for science. Most labs build their own software tools, and they tend to be bad at it. I think there's huge potential for craft-oriented toolmakers to build the tooling needed in computational science, if we can design the right incentive structures for it. I'm working on ways to make that happen, so if you're interested in devtools-for-science then my DMs are open! I've been building some tools over the years, in various stages of abandonment. • Labfile & Lab: a declarative file format and orchestrator that resolves experiment dependencies and runs them in-order (i.e. Terraform for science). • Logis: a Python decorator that turns your git commit history into a queryable scientific log. • Cyantic: a Pydantic preprocessor that makes it easy to build (eg.) Tensors from parameterisations. • Anot: a tool that extracts structured data from source code comments. • Codeline: run context-aware commands from source code comments. • git plan: write your git commit message in-advance.