The Wheel Reinvention Jam is a one-week jam to build software from scratch.
Inspired by Handmade Hero, we believe in the power of reinventing the wheel. It's an opportunity to grow as a programmer, to question long-held assumptions, and to discover how simple and effective software can be. The Wheel Reinvention Jam is a chance to start your own Handmade project and see where it takes you.
Reinventing the wheel is in the DNA of the Handmade community. First of all, it's fun! But there are so many reasons why reinventing the wheel is important, not just for individual programmers, but for the whole software industry:
Every important software project begins as a simple prototype. So take the opportunity to start something new!
[post-jam]: here's a rough demo of rendering the control flow constructs in the executed assembly code. it's very broken and very complicated right now, but it's good enough for a demo screenshot and for me to finally call the jam project done :P &asm-trace
&vxt
Did a write-up on the blog about my entry for the jam.
https://unlitart.com/Blog/Wheel_Reinvention_Jam_2025_Procedural_Tree_Growth_Simulator/
[post-jam]: rewrote the control flow analyzer. each structure now has a unique header (a basic block) that dominates everything inside the structure, which turned out to be a useful invariant. &asm-trace
At the end of the jam &adaw can export audio files that you create. Didn't get in everything I wanted, but this does feel good
Here's a 10-minute video demo showing off all the features of &flowshell as they stand at the conclusion of the jam! I had a ton of fun with this project, I think it turned out quite well, and I think I may have to continue adding more nodes. And undo and redo. And any kind of serialization. https://www.youtube.com/watch?v=VyqxUs1mUng
&pajama day 7 bonus content - No new functionality, but I did a small visual refresh with the last remaining minutes of the jam!
&pajama day 7 - That's a wrap! I got multi-track recording and playback working, complete with independent per-track instruments. There's still a ways to go before this is useful - I'd want midi export and at least some basic track editing - but it was a lot of fun to see how much progress I was able to make in a week. I'm also happy with my project choice for this jam, the immediate gratification of being able to (kind of) make music at increasing levels of sophistication kept me motivated. I'm certainly hoping to keep the momentum going and work on it further in the coming weeks.
&rolledit Ok, I'm toast at this point, going to call it.
I almost made a video editor in a week :P
No UI for adding/removing clips for the timeline unfortunately, but it takes a list of filepaths from the command line to load into the editor, and it can manage to create a valid mp4 out of two big_buck_bunny.mov files stuck together.
I've got a lot of video-editor pieces, but there's some lurking pts/dts bugs kicking around that stop this from being usefully general purpose.
It's jam quality code, can't claim it's worth using for anything other than a quick stare, but I had a lot of fun working on it.
https://github.com/colrdavidson/rolledit
&vxt
It's 4am and I'm calling it for the jam!
This project turned out way more challenging than expected, and in the end I had to resort to ad-hoc solutions to get growth that looks somewhat tree like.
I realise that the proper system will have to involve some skeletal physics simulation of sorts. The paths branches take greatly depends on surrounding factors, such as access to sunlight, and proximity to other branches.
I definitely want to return to this project at some point. The foundation is in place but more research will be required.
days 4-6.83: identifying control flow constructs was way, way, way harder than expected. i'm finally getting somewhere with the analysis. but i'm not even close to using this data to improve the &asm-trace interface.
