Potato Arena combat Madness

Eighty potatoes, two joints each

Hey all! This week was mostly a design doc and an asset problem, so that's what I've got for you.

So I started laying out a game I'd been calling Potato Arena. Clash Royale shaped, built for Lens Studio, simultaneous multiplayer, Potatoes on both sides. I renamed it Potato Arena this week, and that rename is actually the first thing I want to talk about.

If you build a two-lane tower game where you play a hand of cards, people are going to tell you exactly what you built. I didn't want to spend a year reading "this is just Clash Royale" under everything I post, and they'd have a point. So I went looking for a version that only works with my characters in it.

What I landed on is a kitchen. Each player's Potato sits in a fry basket, and you win by sinking the other guy's basket into the deep fryer. The arena is a fast food kitchen interior, stainless steel counter tops with three bridges running between them. And instead of a hand of cards you pull from a potato sack.

The reason I like it is that the stakes come free. Nobody has to explain to you why a potato does not want to go in the fryer!

The part that's actually hard

Here's the problem I spent most of the week on.

I have a lot of Potatoes. Eighty-plus variants built up over the years, and I keep a spreadsheet with metrics on all of them. For a game that wants a roster, that's an absurdly good place to start. Most people building something like this are drawing characters. I have a closet full.

Except every one of those models is an FBX that's rigged and not animated. And the rigs are about as simple as a rig can get, a Head joint and a Body joint, which was completely correct for the job they were built to do, which was sit on somebody's face and bob around a little. Nobody needed a potato to walk.

The joint names aren't consistent either. Some of these models are years apart, built in different moods, and past me was not thinking about this particular day. So animating eighty characters one at a time is not a thing that's going to happen

The plan I'm going with is one master set of motions on one clean rig, and then reuse it across everybody. Idle, walk, attack, hit react, die. Every variant gets the same skeleton structure with the same joint names, and the animation set gets pointed at whichever potato is standing on the counter top. Two joints sounds like nothing to animate with, and it mostly is, but squash and lean and a little bounce timing will carry a character further than people expect. These are potatoes. They don't have elbows to worry about.

The unglamorous half of that plan is opening a lot of old files and renaming joints by hand until they all match. A few evenings of work with nothing to show at the end of it, and no clever way around it that I've found.

I'm still stuck on the art style. Clash Royale reads as a painted illustration that happens to be three dimensional, and I don't have that yet. My potatoes look like 3D models of potatoes. I think the answer is in the shading rather than the models, so that's what I'm digging into next, and if any of you have done this look in Lens Studio I'd love to hear about it.

One other decision worth writing down. I'm building the single player AI opponent before anything fancy, because a brand new lens does not have a matchmaking pool. Two people have to want to play at the same second, and on day one that isn't happening. So the AI isn't a fallback mode I add later, it's the thing that has to be good first.

On the platform side

There was a piece going around on Friday the 28th about how the glasses platforms are split right now, and it put something I'd been feeling into clearer terms. Snap has the whole publishing pipeline ready, developer portal, Lens Studio 5.24, Commerce Kit set to turn on in-Lens purchases and subscriptions at launch, and the glasses are barely out in the world yet. Meta has Ray-Ban Display sitting on store shelves with no public way to submit an app to it. Google has neither one.

I know which problem I'd rather have. I can build against a finished store and wait for devices. I can't do much with devices I can't publish to.

The Commerce Kit part is the one I keep thinking about. I've been running my own payment rail for years because there wasn't one. Potato Cam subscriptions live on Patreon, and I moved part of that over to Gumroad recently. It works, but it's me holding together billing and access and a list of who's current, which is not what I'm good at. Purchases and subscriptions inside the lens itself is a different business than the one I've been running.

The Specs First Look demos also kick off this week and run into November, so a lot of people are about to touch these for the first time. I'm curious what they say when they take them off

Be good little potatoes out there (or be bad ones, either way!) I'll talk to you again soon!