Personal site · projects and notes
Known online as odxnoj, or simply odx.
A project to bring the Firefox 1.0 codebase forward for modern systems. The target is Windows, macOS, and Linux. The work so far has focused on getting the old source to build and resolving the missing pieces needed to launch a browser, not just an executable that prints a version string.
An experimental Buildroot-based Linux system running on an ESP32-S3 with 16 MB flash and 8 MB PSRAM. It uses a NOMMU userspace, so normal Linux assumptions do not always apply. I have been working with SSH, the shell, networking, small command-line tools, and hosting this page from the board.
A personal assistant and hardware experiment built around a laptop and an ESP32-S3. The idea is to connect a local language model with speech, sensors, and eventually a physical interface. Speech recognition and hardware integration are part of the ongoing work.
I experiment with running and adapting small LLMs on limited hardware, including a phone. I am interested in the trade-offs between context, memory, inference speed, prompting, and how well a small model can reason through a technical problem.
Linux internals, memory models, init systems, embedded userspace, Buildroot, and what changes when a system has no MMU. I am especially interested in small systems that can be understood by inspecting how their parts fit together.
Browser internals, old codebases, compatibility problems, and the work required to keep software from another era running on current operating systems. REWIND is where I explore this directly.
I am learning C and working toward a stronger understanding of memory, processes, interfaces, and low-level debugging. On the hardware side, I use the ESP32-S3 for small experiments with networking and peripherals.
Quantum mechanics, the physical basis of consciousness, and the difference between describing a mechanism and explaining the experience it produces. I am still building the mathematical background to study these properly.
I am an egoist, and Max Stirner is a major influence on how I think about the individual and fixed ideas. I agree with Stirner on everything except his views on religion.
I am interested in the point where an idea stops being something a person uses and becomes something that claims authority over them. A rule does not justify itself by being old, popular, or described as sacred. It still has to be examined.
For me, self-interest is not automatically the same as immediate pleasure. Understanding what serves me can require patience, learning, and rejecting an impulse that feels good in the moment.
I am a misanthrope largely because I see double standards as widespread. People can demand honesty, fairness, or loyalty from others while excusing the same behaviour in themselves or their own group. The principle changes depending on who benefits.
This overlaps with the idea of moral hypocrisy: presenting a moral standard while applying it selectively. Related concepts include motivated reasoning and in-group bias, though they describe different mechanisms.
My view is that this pattern is common enough to shape how I judge humanity. That is my judgement based on observation, not a claim that every person behaves this way.
I am still working out how best to define self-interest, what would count as strong evidence about human behaviour, and how much any general view of humanity can survive exceptions. These are questions to investigate, not conclusions I have already solved.
Could a complete account of brain activity explain why experience feels like something from the inside?
If decisions arise from prior causes, what exactly do we mean when we say a person could have chosen otherwise?
How can I distinguish what I actually want from habits, fear, social pressure, or a short-term impulse?
What evidence is enough to move from observing repeated hypocrisy to making a broad claim about human behaviour?
Which limitations come from model size, which come from training, and which can be improved through tools, memory, or a better inference setup?
What is the smallest useful computing system that can still run the tools I care about, and where do its architectural limits begin to matter?
Egoism — Stanford Encyclopedia of Philosophy
A reference on psychological, ethical, and rational egoism, with arguments and objections.
The Ego and His Own — Max Stirner
Primary text, available in several downloadable formats.
“Moral Hypocrisy and Acting for Reasons” — Maureen Sie
A philosophical paper on moral claims, action, and the reasons behind them.
Building a small Linux system, configuring packages, toolchains, and the root filesystem.
ESP-IDF documentation for ESP32-S3
Hardware, peripherals, networking, and development tools for the board.
Kernel interfaces and subsystem documentation.
The original codebase that REWIND is based on.
Contemporary release information and a view of the browser's original scope.
OpenStax University Physics, Volume 3
A free textbook covering optics, relativity, and introductory quantum physics. The mathematical prerequisites matter, so I am approaching it step by step.
Expanded the stated project goal to Windows, macOS, and Linux.
Expanded the site to cover embedded Linux, browser porting, and small-LLM experiments alongside technical notes and reading.
I am interested in systems that are small enough to inspect and complicated enough to teach me something. That has led me to Linux on an ESP32-S3, an attempt to revive an old browser codebase, C, embedded hardware, and experiments with language models on limited memory.
Outside programming, I read books and study physics and consciousness. I tend to follow a question across fields instead of keeping technical work and philosophy in separate boxes.