Floaty in 2026

Last updated on 19th August 2025

Some time ago, I created a JS port of Pico-8. I challenged myself to host a workshop (for work people) where I taught non-coders how to make a game. Games are a powerful tool when teaching programming, because of their rapid iteration cycles, fun outcomes, and a general acceptances of newcomers.

The trouble was, there are no easy ways to teach JS game dev to newcomers. I go into this in one of my talks, but the TL;DR is that there’s just too much to do before you make that first thing jump, speak, or shoot.

I was dabbling in Pico-8, at the time, but could find no JS port to use for the workshop. So, I made one. I followed a YouTube playlist for making a popular type of game. As the presenter used some new Pico-8 feature, I added that same feature to my port. I stopped when he got to music, because I am awful at that and don’t enjoy it at all.

That’s also, consequently, when I stopped a lot of development on Floaty.

I was reflecting this week, and remembered all the things I’m proud of building in the last few quiet (as pertains to social media and conference travel) years. One of those things was Floaty. I want to do more with it. On the one hand, I think it’s important to remember where it has come from. It is a Pico-8 port. This is undeniable when looking at them side-by-side.

On the other hand, I like how Pico-8 describes itself as a virtual console. It’s not alone in this. I want Floaty to stand apart more. I want to leave the parts I don’t enjoy on the cutting room floor. Embrace not making custom sounds or music in-engine (or first-party tools). Sure, I can explore browser-based sound generation, but why not support wav/mp3/ogg sound and music files as first-class citizens, and allow others to shape what first-party support happens for browser-based sound generation?

I could drop the heavy weight of ToneJS. The heavy weight of feeling like I have to make all these tools (i’m not enjoying) to fulfil the process of making music (I’m not enjoying); just so I can get as close to Pico-8 as possible. Isn’t it enough that someone could follow 80% of that YouTube playlist while writing the game in a completely different engine?

I am still heavily invested in the idea of building sprite and map features, but I want to emulate Pico-8 more in that I want to build Floaty games that export code Floaty could use. Not the Livewire monolith I’ve been struggling to build and maintain. Just simple tools to get you 80% of the way to making something that feels like a Pico-8 game, but in a language that’s easier to teach and more useful to know.

Not that I don’t like Lua. It’s just not gonna make you any money.