I'm bootstrapping a malleable operating and development environment called Tractor Studio, intended to become a "Photoshop of software." Towards this goal, I'm working on several powerful software primitives that also incidentally build out Tractor Studio. I refer to these primitives collectively as the Tractor system. Currently released and open source are Apptron and Treehouse. Two projects that will be out soon are Manifold and Tractor Workbench.
Apptron is a primitive for working with native desktop APIs and building cross platform webview apps, similar to Electron but much lighter and language agnostic. Apptron is an example of a high-leverage software primitive.
Treehouse is a toolkit for building powerful information tools like Notion and Obsidian. As a toolkit, it comes with a fully usable "note-taking" frontend that you can extend or customize and then "bring your own backend". This project is intended to validate lightweight, generic frontends that should exist for common types of apps, and provide a framework for future generic frontends.
The core primitive of Tractor is called Manifold, which is an embeddable runtime for assembling software from disparate components into a common data structure and API. This can be used directly by operators and developers, or be used with generic or specialized frontend interfaces for end-users.
A prototype of Tractor Studio is being developed to be open sourced soon called Tractor Workbench, which is an extension to VS Code for working with Manifold. Workbench is intended to inform and bootstrap Tractor Studio, which itself is to be built through Manifold.
I'm always looking for help and contributors to the active open source projects (Apptron, Treehouse) in any capacity (dev, docs, maintainers, marketing, etc), but I'd really like to find an open source hacker to help me start a new R&D project around universal API schemas. Not just web APIs, but libraries as well.
I want to build an open repository and toolchain for building out and collectively maintaining schemas for any API. These schemas could be used for code generation, documentation, interactive consoles, but I'm interested in using them to data-drive any system with user facing external integrations or "connectors" found in many low-code/no-code offerings. This is to help reduce the code duplication across these products and projects, and make these kinds of integrations more accessible. It would also serve Tractor Studio.
I've done lots of work with OpenAPI, I've used Electron's API schema to re-expose the API in other languages (early Apptron research), I've built out a toolchain (github.com/progrium/macschema) to parse Apple documentation to produce schemas used to generate Go bindings (github.com/progrium/macdriver), and a few other preliminary experiments in this direction.
There is probably a startup in this alone, but being an open repository and having crowdsource-oriented tooling is a big part of the vision. Let's collaborate and see where it goes!
Jeff is a rogue software engineer and veteran open source hacker. He is a founder of SuperHappyDevHouse and Hacker Dojo, and helped create Docker, Twilio, and TIGSource. He also pioneered webhooks and early ideas around serverless.
Mission: Enable people to build and integrate arbitrary software systems from high leverage primitives through the "Photoshop of software."
Interests: Obviously tools for tool-makers, but ultimately tools to unlock creativity in all directions. I'm also into music (playing/writing/producing), films and filmmaking (yet to be explored), and indie game development (completed several small projects). I'm big into history as well, specifically of technology and computers (therefore retro computing), and have a semi-obsession with systems theory and systems thinking.
- In the 2000s I helped start legendary, influential hacker parties in Silicon Valley called SuperHappyDevHouse, and later the hackerspace Hacker Dojo.
- I coined and pioneered webhooks with an eye towards the no-code/low-code world it would enable.
- I consulted on the design and helped prototype Docker, then developed a number of tools in that ecosystem under the banner Glider Labs.
- In general, I've developed many open source tools and libraries, some pretty influential like Dokku, logspout, registrator, localtunnel, JWT, RelMeAuth, and RequestBin.
- I've worked and consulted on many different kinds of software projects with lots of organizations, including NASA Ames, Twilio, Netflix, Hashicorp, American Express, DigitalOcean, and Docker.