Home

Welcome to the TR4QT wiki! TR4QT is an experiment built on a long lineage of contest logging software. TR4W (an open source project on GitHub) stands on the shoulders of TR4W, and TR4QT continues that evolution in a modern, cross‑platform Qt/C++ application.

Over Christmas 2025, inspired by Mike AI5QK’s work using AI tools to generate cross‑platform Qt/C++ code to control the Elecraft K4, I started porting TR4W to TR4QT. The goal is to re‑implement many of the features of TR4W (and TRLOG before it) purely by crafting careful prompts and iterating between an AI coding agent and my own 40 years of software development experience. The AI does not produce good code on its own; that’s simply the reality of today’s coding agents. However, when you enforce clear architectural principles, strict design constraints, and a strong testing discipline, the resulting code becomes surprisingly good and well structured.

Left to itself, the AI tends to create huge “god classes” and single giant methods, as if it were writing a monolithic COBOL program. With guidance toward polymorphism, proper factories, elimination of magic numbers, and clean abstractions, the codebase becomes maintainable and robust. In this project, the agent has written 100% of the code; I have not authored any of the implementation directly. Even so, the result is already feature‑rich and performant.

This is still very much a work in progress. If you use TR4QT for a serious contest effort today, you should do so with full awareness of the risks. ADIF and Cabrillo export are still under active development, as is CW keying support, so expect rough edges. You can explore and run TR4QT on multiple platforms: macOS, Windows, Linux (at least Ubuntu), and Raspberry Pi. Radio control is network‑based, with direct support for Icom rigs (so far tested with the IC‑9700 and IC‑7610), and Hamlib support for a wide range of other radios.

Check the CNANGELOG.md file for details on what has been added recently, and be sure to read the main README.md in the repository.

This project is also a vehicle for learning more about AI‑assisted “vibe coding” and GitHub Actions for automated builds in a continuous integration (CI) pipeline. If I can help move TR into a modern language and architecture along the way, so much the better. Work on TR4W continues in parallel; the AI has contributed there as well, though not all changes have been merged to  main  yet.