#wildmender

And then there's fucking Wildmender. A game where you are a single human child in a world of endless wasteland and death, where the only other things are ghosts who remember a halcyon era and the hubris that ended it, wraiths which are consumed by their own greed and destruction of the land for their cursed immortality, and a couple god statues. The entire map is just ceaseless grief, filled with the literal dessicated remains of all the biodiversity that came before the countless disasters. And it's a big fucking map.

And then...the game gives you a shovel and a sickle and a mirror that shows the wraiths what twisted reflections they've become.

And the game says, "The entire world is waiting to be better, and the only way to do that is by doing it yourself, long and hard and hopeless as it seems."

I cannot emphasize enough how overwhelming the task you're handed. There is not a single speck of life left in the world. You are given a shovel and a water bottle and just...expected to do something about it. To look at the literal endless wastes and think you can heal it.

This is what Wildmender cherishes that Terra Nil denies: This is an impossible task for you alone. But it has to be done...and you can actually do it. The way you can turn sand into soil and dig irrigation channels is beautiful. Every single scrap of land that you reclaim is something you had to do on purpose. You had to do it yourself. You had to actively choose how to do it.

And the game makes the reward of even just getting a bit more water into the sand feel like victory. Your starting oasis turns from soil into lush and beautiful meadows--sure, technically instantaneously by doing magic on a specific type of plant. But it took me 4-5 hours before I got there. You have to travel so far into the desert to learn how to grow grass again, and then you realize that this endless hostile wasteland is a fraction of the map you're given. And you look at this sudden profusion of meadowy grassland surrounding your spring and despite how sudden it feels you remember how big the world is. You made more progress in a minute than you did in 5 hours and it's not even a speck on the map. How the fuck is this gonna happen?

And the answer is by accepting that it's going to take a long fucking time and a lot of hard work.

Anyways here's some Wildmender fanart of Naia Mother Of Rivers

she is my mom and also kitty

So I've been playing this game called Wildmender where you wake up in a desert and have to build an oasis (and unblock or cleanse other springs, plant vegetation and help regrow the desert- all while being attacked by wraiths and trying to purge evil from Gods shrines so they can lend you their powers- and lots of other stuff, its great)

(pictured above: Wraith- sometimes they just spy on you and don't attack and lemme tell you i've jumped out of my skin more times than I can remember from turning around to spot one, just silent and still, staring at me from the other end of the garden. Much scarier than being chased by like 10 of them hurling fireballs)

But mostly it's just an incredibly beautiful game

Over the last 4 years of work, we've created a gardening survival game in a desert world that let people create massive and complex oasis where each plant is alive. At the heart of a system-driven, procedurally generated ecology is a water simulation and terraforming system, and this post is to share a bit of how we built it.



We knew early on that the game would include some level of soil and water simulation, just as an outgrowth of wanting to simulate an ecosystem. The early builds used a simple set of flags for soil, which plants could respond to or modify, and had water present only as static objects. Each tile of soil (a 1x1 “meter” square of the game world) could be rock or sand, and have a certain level of fertility or toxicity. Establishing this early put limits on how much we could scale the world, since we knew we needed to store a certain amount of data for each tile.



Since the terrain was already being procedurally generated, letting players shape it to customize their garden was a pretty natural thing to add. The water simulation was added at first in response to this - flat, static bodies of water could easily create very strange results if we let the player dig terrain out from under them. Another neat benefit of this simulation was that it made water a fixed-sum resource - anything the player took out of the ground for their own use wasn’t available for plants, and vice versa. This really resonated well with the whole concept of desert survival and water as a critical resource.



The water simulation at its core is a grid-based solution. Tiles with a higher water level spread it to adjacent tiles in discrete steps. We broke the world up into “simulation cells” (of 32 by 32 tiles each) which let us break things like the water simulation into smaller chunks that we could compute in the background without interrupting the player. The amount of water in each tile is then combined with the height of the underlying terrain to create a water mesh for each simulation cell. Later on, this same simulation cell concept helped us with various optimizations - we could turn off all the water calculations and extra data on cells that didn’t have any water, which is most of the world.



Early on, we were mostly concerned with just communicating what was happening with simple blocks of color - but once the basic simulation worked, we needed to decide how the water should look for the final game. Given the stylized look we were building for the rest of the game, we decided the water should be similarly stylized - the blue-and-white colors made this critical resource stand out to the player more than a more muted, natural, transparent appearance did. White “foam” was added to create clear edges for any body of water (through a combination of screen depth, height above the terrain, and noise.)

We tweaked the water rendering repeatedly over the rest of the project, adding features to the simulation and a custom water shader that relied on data the simulation provided. Flowing water was indicated with animated textures based on the height difference, using a texturing technique called flowmaps. Different colors would indicate clean or toxic water. Purely aesthetic touches like cleaning up the edges of bodies of water, smooth animation of the water mesh, and GPU tessellation on high-end machines got added over time, as well.

The “simulation cell” concept also came into play as we built up the idea of biome transformations. Under the hood, living plants contribute “biomass” to nearby cells, while other factors like wind erosion remove biomass - but if enough accumulates, the cell changes to a new biome, which typically makes survival easier for both plants and players. This system provided a good, organic feel, and it fulfilled one of our main goals of making the player’s home garden an inherently safe and sheltered place - but the way it worked was pretty opaque to players. Various tricks of terrain texturing helped address this, showing changes around plants that were creating a biome transition before that transition actually happened.



As we fleshed out the rest of the game, we started adding new ways to interact with the system we already had. The spade and its upgrades had existed from fairly early, but playtesting revealed a big demand for tools that would help shape the garden at a larger scale. The Earthwright’s Chisel, which allowed the players to manipulate terrain on a larger scale such as digging an entire trench at once, attempted to do this in a way that was both powerful and imprecise, so it didn’t completely overshadow the spade.

We also extended the original biome system with the concept of Mothers and distinct habitats. Mothers gave players more direct control over how their garden developed, in a way that was visibly transformative and rewarding. Giving the ability to create Mothers as a reward for each temple tied back into our basic exploration and growth loops. And while the “advanced” biomes are still generally all better than the base desert, specializing plants to prefer their specific habitats made choosing which biome to create a more meaningful choice.