Memoranda

On this page, I will publish notes, essays, reports and other writings that have been produced by myself. The contents can be changed without notice or any form of reporting. Comments can be sent to me personally using the contact form.

Want to access your TP-Link TL-WA901ND over HTTPS using a local domain and an nginx reverse proxy? Turns out it’s trickier than expected — thanks to some hardcoded quirks in the web interface. In this memo, I walk through the setup I used (Raspberry Pi, Pi-hole, custom domain), explain the issues I ran into, and share the final working config. If you’re running a homelab and want cleaner, secure access to your AP, this might save you some trial and error.

The Libertinus Keyboard font offers a clever way to display keyboard key symbols using ligatures — special character combinations that automatically turn text into a key icon, like Shift. While implementing Libertinus on this site, I found little documentation, so I’ve written a short memo explaining how it works. It includes examples, quirks, and a full list of available key glyphs extracted from the SFD source. A must-read if you’re working with Libertinus, typography, or accessible keyboard icons in web projects.

A concise overview of the daily-use machines I work with, including detailed hardware and system specs gathered via inxi. Machines are named after fruit-bearing plants following a personal scheme, with specs for devices like the Lenovo ThinkPad L380 Yoga (ananas) and a desktop with Intel Core i7 and NVIDIA GTX 1080 (durian). This list helps clarify the environments where I test software, noting that functionality on other machines isn’t guaranteed.

A detailed overview of Genetic Algorithm (GA) methods based on a literature review, illustrating key concepts such as selection, crossover, mutation, replacement, and termination. Using a custom visual language based on RGB fitness values, the memo explains common GA techniques like Roulette Wheel Selection, Tournament Selection, Single- and Multi-Point Crossover, and various mutation strategies. Mathematical formulas and clear diagrams accompany the explanations, providing a solid foundation for understanding and applying GAs in optimization problems.

I discuss the importance of connected components in image processing and computer vision, focusing on their role in image segmentation and analysis. In this text, I explore how connected components, especially within Max-Trees, can be used to derive meaningful information from images by computing various attributes. I also provide GNU Octave implementations of these attributes to facilitate practical application.

I introduce the concept of Max-Trees, an advanced technique in image processing that structures image data beyond simple pixel sequences. I explain the process of constructing a Max-Tree and its importance in image analysis, including a practical GNU Octave implementation. This explanation builds on research from my 2021 master’s thesis on detecting astronomical objects using machine learning.

This website collects statistical data in order to improve your user experience.

Privacy Statement