Crafting a Non-Violent Post-Apocalyptic Deckbuilder Inspired by Japanese Woodblock Prints
Max Shawabkeh shares what inspired Worlds Upon The Wind, explains how he created the visual style of the game based on Japanese woodblock prints of the 1600s, and comments on how he developed the story and set up the game mechanics that would feel right for it.
Introduction
After quitting my job, I decided to try out Godot by building a tiny game to practice Japanese kanji characters. I managed to get a basic version within a day, so I decided to expand it with a game mechanic, and after brainstorming for a while, I landed on one about matching elements to slots. Since I've always preferred making non-violent games, a city-building theme was an easy choice.
At the time, I was also reading along with my girlfriend's Japanese Literature course, so between that and the cards being Japanese characters, a Japan-inspired setting seemed like a great idea. I considered using art inspired by Japanese ukiyo-e woodblock prints, and since ukiyo-e literally means "pictures of the floating world," I decided to set it on floating islands. With those key decisions made, I needed a reason for the islands to be floating, and an ecological disaster seemed timely given the current state of the world.
Over time, this evolved into the game as it is today, a story of overcoming grief for a shattered world spanning hundreds of years.
Creating Worlds Upon The Wind
Ultimately, I made a game that I would enjoy. It has the kind of gameplay you would find in a deckbuilder, but instead of fighting and power, it is about restoration and maintaining balance. I think people who appreciate careful planning would adore the gameplay most, while folks who like a relaxing, contemplative experience would enjoy the world and the presentation, with its classical art and calm music.
I was going for a peaceful, serene atmosphere, but players have called it everything you could imagine, from 'relaxing' to 'zen' to 'insanely intense.' This is partly because it's designed to let the player decide how much of a challenge they want to take on. You could finish the whole story without ever going beyond a dozen cities per run, or you could spend 5 hours each run filling up the entire map and facing extreme challenges.
Choosing the Software
The game is built using the Godot engine. For my last few projects, I had mostly used Unreal Engine, and moving to Godot was a breath of fresh air. The engine is clean, consistent, and an absolute pleasure to use.
The vast majority of the game is built using GDScript, which makes it a breeze to add features, create new content, and debug. Of course, it isn't as performant as a compiled language, so for example, once I was done prototyping the map generation system, I ported it to C++. Godot's GDExtension system made this very easy; once that one system was ported, the rest of the game could continue using it without requiring any changes.
Over the 18 months of development, I only encountered a few engine bugs, and the Godot developers went out of their way to fix them or accept my fixes. There was only one case where they couldn't do it in time for the game's release, so I had to write my own workaround, but that's still a great experience compared with UE.
Game Mechanics
The game's core mechanic is very simple: glyphs are cards that have essence icons, and at each stage, there are essence slots that can be filled with any glyph with a matching essence. These slots are used everywhere in the game: it's what you use to set up a farm, connect a city to the capital, counter a natural disaster, or choose an action during a scouting mission.
In addition to standard deckbuilding mechanics, what makes it interesting is the many layers built on top of this. Instead of using glyphs for essences, you can activate their abilities like recycling glyphs, copying them, gaining resources, and so on.
Then there are hundreds of event vignettes which can grant relics. These either change some rule in the game or trigger off various actions, allowing players to set up chain reactions. Then there are hauntings, obstacles that cause negative effects when certain actions are taken. And then there are various animal companions, which provide active abilities that can completely change the player's playstyle, and more.
On top of this are about a hundred metaprogression skills that let the player access new mechanics and give them more choices in how to customize their strategies.
Visual Style
I am only an amateur artist myself, so I knew making all of the art from scratch would take me years. Early on, long before the game was announced, I experimented with AI art, but given all of the issues surrounding it, I decided to try remixing public-domain historical art instead. This was very slow. I spent much of the summer of 2025 researching Japanese woodblock prints and picture scrolls and then learning to mimic the style in Photoshop.
Eventually, I ended up using historical art in 2 ways:
- Static illustrations use whole woodblock prints. Most are touched up to match the game's color scheme, and some are edited in Photoshop, for example, to remove weapons or to mix the background of one print with the foreground of another. The majority of my time with these was spent using the Clone Stamp tool.
- Dynamic scenes like the game's procedurally generated maps and the evolving hub area use a mixture of pieces cut out of historical artworks with shader-based coloring and subtle animations, plus there are some parts I drew in Photoshop that mimic the traditional style.
The rest of the art, such as the UI, relic icons, animal companions, and so on, I ended up drawing from scratch.
Here's an example of how the map was made. First, there's the original woodblock print, a travel guide by Hishikawa Moronobu from the 1600s. Here's a sample page from it:
This was cut into separate pieces by K. M. Alexander, and his work was what inspired me to take a similar approach:
I took these pieces and colored them in to indicate material layers:
Then, in the game, a fragment shader is used to apply colors and textures to these sprites. Things like trees and rivers are then animated with a vertex shader to simulate wind or rushing water.
This gave me the basic pieces for the map, but there are 200+ unique structures in the game, so I ended up spending a few months drawing new sprites in a similar style from scratch. Here's a sample:
Game Story
The game's plot is quite simple: the world has shattered, and now society has to be rebuilt from scratch. The survivors face various challenges and try various solutions, eventually reaching a conclusion on how to deal with the consequences of their society's mistakes.
Although simple, the story does not play out how a typical Western post-apocalyptic narrative would, and ends up addressing several issues relevant to today's world, such as the nature of leadership, belief, environmental stewardship, and cultural identity.
I think what makes the whole thing effective is that the world is developed through hundreds of tiny stories. You might have to decide how a festival is held, how a disagreement is resolved, or how unexplained phenomena are approached. Then, as you settle the shards of this broken world, they start to develop their own cultures and histories. All of these elements are quite consistent and paint an immersive, grounded world for the player to look after.
Conclusion
What word of advice do I have for beginners? This is a much harder question today than just a few years ago. Back then, the main challenge for an indie developer was getting the game noticed and delivered into the hands of players. Even great games had only a small chance of making enough profit for development to be sustainable.
That challenge remains, of course, but I am afraid traditional game development skills are not going to remain marketable for long with the rise of AI. Already we see Steam filled with vibe-coded, "vibe-arted" games. So far they are mostly pretty bad, but the technology is advancing quickly, and I wouldn't be surprised to see some AI-based games become hits in a year or two. As much as I hate to recommend it, if your goal is to make a living doing this, you will probably have to learn to adopt AI tools.
On the other hand, if, like me, you want to make games for the love of the craft, and you aren't heavily relying on the revenue, the advice is a lot simpler. If you are just starting out, pick an existing engine and finish projects while aggressively trimming scope. A finished, fully playable game, no matter how small or janky, will teach you more than two years' worth of tutorials. Try the smallest game you can imagine. Then one slightly bigger. Then an even bigger one. And eventually you'll end up with something worth players' attention.
If you are preparing to make a larger game with a real audience, get it in front of playtesters as soon as possible. I had friends trying out Worlds Upon The Wind within two weeks of starting, and that gave me a year and a half of player feedback, which was invaluable for getting the game to the level of polish it has today, which is something that gets praised consistently in player reviews.
