
Quality or collapse in the age of LLM coding
LLMs are already a part of coding whether you like it or not. Still undecided is its trajectory: towards better software or towards accelerated production of garbage.

LLMs are already a part of coding whether you like it or not. Still undecided is its trajectory: towards better software or towards accelerated production of garbage.

How I used AnimationMode in Unity to tell the editor that I’m editing values and have them revert when the preview mode stops.

An exploration of how interpreters processing code introduces bottlenecks that compiled code does not, and some techniques to alleviate those issues.

Explore how React’s useState hook works under the hood and understand how state is tracked across renders — a key to writing correct and predictable code.

Our id-source system is a universal system for mapping internal IDs to readable, localized names across frontend and backend. It powers dropdowns, labels, and search with features like progressive loading, custom formatting, and metadata support—solving a core challenge in scalable app development.

A method to provide common data analysis and exploration, such as lists, filtering, searching, analysis, editing and charting, across a variety of heterogenous tables, without code duplication.

An implementation of a discrete item probability queue in C#. I wrote this to implement rarity based distribution of items in my game Rabanaz.

Designing a data manager that coordinates network requests to avoid duplicate loading

There are too many ways to represent unknown or missing data and this can lead to defects.

My hopes for the current generation of language modelling and machine learning