Selection of various technical things I'm working on. Some more finished than others, a few case studies.
gdoc_sections
Closed Source
Configurable library + CLI tool to parse sections out of Google Docs. Can convert Docs' internal JSON format into plain text, web-ready HTML, or HTML that can be inserted into the clipboard to paste into Docs while maintaining formatting (header levels etc.).
bell
Prototype under active development. Define resources, track stocks/inventory, set up production methods to convert resources to other resources. Place fixed work orders or ad hoc transactions, set production goals to produce resources by automatically placing work orders. View past statistics and construct future plans.
sourcery
Closed Source
Uses gdocs_sections to extract a set of sections representing RPG rules to a SQLite database so they can be tracked and associated with characters. Syncs with character documents with out-of-sync checking and reconciliation. Draws rules options from the master rules document while respecting prerequisites on individual characters. Faithfully displays rules formatting in the dashboard as in docs, and copies to clipboard in a format that pastes perfectly into Docs.
mcmod
Probably my largest and most complete personal project. Lots of interesting subsystems, APIs, etc., coupled to a large codebase that has changed a lot since I started. In development since ~2019.
mt
Closed Source
Apply tags to locally-stored media for categorisation and ease of searching. Easy-to-use syntax for complex search queries on the database. Easily copy images or file paths. Suggests likely tags with a Bayesian model for faster tagging.
Various Graphics Projects
Some C++ work I did with SDL3_GPU to play around with Vulkan-style rendering close to the GPU.
ECS
A lightweight little Entity-Component System in C, designed to be cache-friendly and data oriented. Used for a few miscellaneous projects.
Mapper
Closed Source
A mapping application where you can incrementally create timelines with a spatial display. Writeup mostly contains notes on a fun problem that I threw at frontier models, which they consistently failed. Thought it was interesting so solved it myself afterwards.
Research
Despite nominally being in mathematical physics, I spent almost all of my time writing code and did comparatively little mathematics. Lots of parallel computing work on HPC systems.