TL;DR: Snowboard Kids 2: Recompiled is available for Linux, Mac and Windows. Following the completion of the Snowboard Kids 2 decompilation , I’ve been focused 1 on getting the recompilation into a good state. Now that the worst bugs have been squashed, I’m pleased to announce the public release of Snowboard Kids 2: Recompiled . This recomp is only possible thanks to support from the N64 Recomp…
I’m very pleased to announce that Snowboard Kids 2 is 100% decompiled! All of the game’s functions have now been implemented in C and compile to assembly that matches the original game. There’s still some occasional __asm__ hackery, 1 and plenty of code needs better names and documentation, but every function now has a matching C implementation. That matters because a matching…
In my previous posts, I described how coding agents could be used to decompile Nintendo 64 games and that one-shot decompilation was very effective . That approach allowed me to make rapid progress on the Snowboard Kids 2 decompilation , with the percentage of matched code quickly growing from around 25% to 58%. After that, progress slowed dramatically, requiring me to significantly alter my…
Recently, I have started using a debugger to better understand the runtime behaviour of Snowboard Kids 2 . Debuggers are useful not only for tracking down crashes, but also for validating assumptions: when a function is called, what its inputs look like, and what effect it has on the game’s state. For example, if we suspect that a particular function loads character data, we can set a…
Recently, I’ve been experimenting with ‘one-shot’ decompilation, leveraging Claude’s headless mode in a continuous loop. 1 The results have been surprisingly positive. In the three weeks since adopting this workflow, I’ve made more progress on Snowboard Kids 2 than in the preceding three months. 2 ‘One-shot’ in this context means that Claude follows the…
Recently, I’ve been working on a matching decompilation of Snowboard Kids 2 , an incredibly underrated racing game for the Nintendo 64. The purpose of this post is to document how coding agents have and haven’t helped with the decompilation process. While much has been written about LLMs, far less has been written about decompilation, 1 so I’m adding another data point. A few lessons…