Volcango Devblog Logo

The Beginning of a "Small" Indie Game Project

Hey it's me, H.! What follows is an edited archive of a devblog I made in a gamedev discord. I parted ways with the discord over their allowance and protection of gamedevs who use AIgen assets in their games. Anyway, it makes more sense to have the Volcango devblog public for everyone to enjoy.

What I've edited

This devblog is neither live nor chronological; some challenges required multiple weeks of work to fully resolve. I grouped paragraphs about the same development challenge to be in the same post. I also rewrote my prose to be easier to parse.

The Volcano Appears

Volcango's initial, early, ugly baby stage. An unpolished illustration of a volcano rim with tropical islands in the background. Orange animals gather around the rim. Scribbles and sketches mark work-in-progress elements such as play cards and upgrades.

Scratch art to figure out the proportion of the animals to the board, and a genuine attempt to get the tropical 'feel' of the game.

Initial blog post circa spring, 2025: Me and my spouse make video games! I do the art/animation and he does the programming. We made two small masters-study type games to get the hang of our craft: (Meowy Catmas 2, apis Volat). After that we fished around for a bigger (but still low-scope) project to stretch our stamina.

I proposed an old papercraft/iPad hybrid game, Volcango. I made this prototype for a game jam back in 2012 with seasoned game dev LoneSpelunker, an art intern, and a team of university students. I felt confident about choosing this project because this game had original mechanics I'd invented and knew were playable since I'd prototyped it and tested it before, even if it was over a decade ago. The scope was small, but bigger than previous projects. Devin decided to go along with my plan. Our goal was to make a game board that feels very alive, inviting, and engaging, even though the gameplay mechanics and rng are brutal. This is a game of 'multiple short plays' and rarely overstays its welcome.

I dubbed the mechanics as ‘Battle Roulette;, where players are distributed animal pieces on a volcanic roulette wheel, then given limited tools to manage their positions on the wheel, all while trying to tactically explode the volcano on everyone else.

At the time I originally wrote this devblog post, we had:

A working prototype
Network connectivity
Cute animals
No papercraft :(
A bright cyan win screen where the dodo bird always wins because we haven't designed any screens beyond the playfield
Utterly confusing turn orders
A few test plays where the game was like, three rounds long, which might have been too brutal...
Hopes and dreams
A lonely game artist who wanted a space to upload progress for fun!

The First Big UX Problem

Players did not understand when, or why, the volcano would erupt. This was a crucial part of the game's experience that we needed to make clear.

Around the Volcango board, there are five types of spaces to land on, four of which annoy one of the island's gods. Each time a god is annoyed, it gets one step closer to erupting the volcano. Some gods (quetzalcoatlus, tyrannosaurus) are easier to annoy than other gods (brachiosaurus, dunkleosteus).

An early version of the volcanic eruption 'countdown' mechanic, represented by long chains
hanging from offscreen. At the end of each chain is a colorful charm: Purple brachiosaurus, blue
dunkleosteus, red tyrannosaurus, and green
quetzalcoatlus.

I started out designing this 'chain' UI that dangled in the corner to indicate when the volcano was going to erupt.

Each time a god got annoyed, its chain would yank up, losing a link. Once the chain lost all of its links, the volcano would erupt. I thought this would be a fun, easy UI, maybe incorporating gentle clicks of stone chains rattling peacefully in the wind.

In early testing, I never once checked the upper left corner to see if the volcano would erupt. Even numerical counters laid on top of the idols didn't help. I was constantly surprised when my guys would die in a big eruption. All of my attention was on the game board, not the sky.

Digital art of four unassuming pillars in the middle of a lava field, each with a statue of a god on top: Purple brachiosaur, blue dunkleosteus, red tyrannosaurus, and green quetzalcoatlus. The pillars are marked with bands of lava to indicate how far down they go when they melt into the lava. They also have the god symbols embedded in their tops. Next, I created a 'pillar' UI, seen here.

What I got right: These pillars are plonked right in the center of the game board. Each time a god gets annoyed, its pillar melts into the lava. When the idol falls into the pool, it transforms into the god, which blows people up. The pillars needed masked with a png mask in Unity to look like they were sinking. Anyway, players are already paying a lot of attention to the volcano, so why not put an important play cue there?

What I got wrong: The idols didn’t look like their symbols on the board and would cause problems for colorblind players. Putting the symbols on the pillars with the idols meant the symbols would have to be a lot smaller than the idols. It was a lot of design elements competing for attention and for such a crucial UI communication, that made me nervous.

I had the feeling that making game actions obvious and apparent to the player was the biggest, meanest hurdle we had, even meaner than sound design, and we had to solve it before I did detail work (like animating the animals).

Another try at the pillars, this time to accommodate Devin's reasonable request that a UI element look more even. I eventually redesigned the pillars to be more consistently-shaped and with a better-fitting pixel mask. They were neither better nor worse than they were before, aesthetically, but they functioned more accurately as UI elements. It took me away from private time with my tyrannosaurus which made me grumpy.

What was next? Oh right, the lava pool.

My first attempt at animating lava looked like a butthole. I'll spare you the screenshot. I was trying to tween different linked meshes in and out and...it just wasn't pretty.

First attempt at lava animation, basically looks like a molten flan bouncing up and down since it lacks the volcanic cup and ridge. I did, however get the hang of weighted mesh and physics constraints. Mostly. Whenever I look at this rig now, it's very delicate and difficult to edit, because it's definitely not set up correctly. I landed on this and was shy about editing it further, so this is how it's going to be forever.

Next lava attempt with very shivery-looking lava bubble animation on top of a bouncing mesh. Next, I experimented with swapping through pngs on a mesh in Spine2D. I’ve always been more comfortable with frame-by-frame animation than with tweening. My first attempt was a too violent-looking.

I knuckled down on my animation basics and focused on ease-ins, ease-outs. I redid every frame from scratch. I got to a stopping point not because the animation itself was completed, but because I wanted to see how it worked on the volcano. I looped my developer in and today he implemented the lava pool into the volcano itself.

Third attempt at lava, this time more relaxed-looking. A slower framerate and more careful in-betweens made this feel much more relaxed, like something you could hang out with...until an eruption happens, of course.

The good news is, the animated lava effect itself was not overwhelming. There were minor things I needed to edit on other objects in the scene (at the time, rock glows were the wrong color and animated not-quite-well, interior of volcano base ought to be purple, other stuff) to make them play nice with the lava pool, but the general gist of 'lava' is working far better than I ever dreamed possible. The game was fun to look at. That was a major goal of mine.

However, I needed to redesign the god pillars in the volcano as a result. I thought I was so close to finishing the volcano itself and being free to move onto individual characters to animate, but pillars kept requiring more work. I wish I had known more about structuring Spine files and getting animations to coordinate with each other across different objects.

I still don't understand how I can simultaneously feel so smart and so stupid at the same time but I guess that's the entirety of human existence, so… enter cosmic, existential shrug here, I guess.

Lava pool with pillars inside, now with a bit of the animation appearing to interact with the pillars, all of which were turned into tyrannosauruses for Cool Testing Purposes. The pillars weren’t quite selling their illusion of sinking into the lava, so I kept trying to get lava ‘collars’ to work on top of them. They never coalesced. What I had to do instead was ‘attach' the lava to the pillars by animating bits of lava inside of the mesh, figured out via conversation that the pillars need to be a separate object. This will be okay, I just have to go into Unity and move the pillars myself until they snap to the mesh!

Post mortem update on that: I did not wind up going into Unity. I coordinated with Devin, who put the pillars into Unity for me. I shouldn’t be in Unity, I will cause problems if I go into Unity. Same reason you don’t put a cat in the dishwasher.

The final result of the lava is what you see in the game itself. I’m so proud of the lava pool idle animation. It was also my first time using weighted meshes and physics constraints in Spine2D.

The pillars still needed one more aesthetic update, though. The symbols each pillar represented were very small, even on desktop, sometimes overwhelmed by the shape of the god statue.

Screenshot of the pillars as they appear in-game, nestled in their lava home. To counter this, I worked the god symbols into the god statues as major, dominant elements. Each god becomes less obvious, but they are better as teasers anyway. The important part that players need to see are the symbols and how they match each other.

Next on Volcango devlog:

Developing a Turn Order UI that made sense to players. Also, the decision to include something stupid that wound up impacting the entire game.