A rotating set of things we're building, exploring, or quietly plotting. Some are client work, some are R&D, some are the kind of side projects that refuse to stay side projects.
We recently launched headius.com ‒ a Bridgetown-powered marketing site for Charles Nutter's JRuby consultancy. Initial launch is live, with the blog migration, a refresh of JRuby.org, and ongoing tooling work still to come. Alongside that, we're quietly building a CMS-style layer that will let non-technical users manage Bridgetown sites without needing to touch a terminal or a pull request.
headius.com Static Sites · Bridgetown · CMS for HumansWe're collaborating with our friends at Magnolia Development on something we can't talk about just yet. What we can say: it plays to both teams' strengths ‒ thoughtful design and accessibility from Magnolia, architectural and backend muscle from us. More news when we can share it.
Unannounced · Accessible by Design · Watch This SpaceSnap is a Logo environment for teaching kids to code in real Ruby ‒ not a block language, not a Ruby-flavoured DSL, but actual Ruby with variables, methods, and loops. The turtle moves, the line draws, and the feedback is immediate. It's been through three implementations (JRuby + SWT, Opal, and DragonRuby), and the story of those attempts is its own lesson in choosing runtimes. Ongoing work, driven by the belief that the next generation of Ruby programmers deserves the same magic we had on a ZX Spectrum.
Education · Ruby · Turtles All the Way DownJava has excellent documentation. JRuby developers have to read it and mentally translate everything to Ruby idioms ‒ `CamelCase` method names to `snake_case`, Java types to Ruby ones, constructor conventions, and so on. JRubyDocs is a tool that converts Javadoc into documentation that looks and reads like Ruby, so you can call `file.read_lines` without first learning that the underlying method is `readAllLines`. A small quality-of-life project with outsized value for anyone working on the JVM in Ruby.
Developer Experience · JRuby · Documentation AlchemyA growing collection of building blocks for running language models locally ‒ embeddings, retrieval, evaluation, and the glue that holds them together. Some of this feeds directly into client work; some is pure R&D; all of it is the substrate for larger projects still taking shape. If you're building with local models and need someone who has already hit the sharp edges, we should talk.
AI · Research & Development · Building Blocks