Posted16 April 202612 screenshots

The six-month update

A grid of nine captures from the game, including snow-covered trees in fog, a lake at sunset, the world map, a campfire on a beach, rain over water and a Brachiosaurus in the rain

It has been about six months since I properly started building Evolve.

Actually, it is probably closer to seven months now, but six-month update sounds better.

Things are going fairly well, so I thought it was worth doing a broader update on where the game actually stands today.

There is now something reasonably close to a playable game.

The world generates. Trees and foliage generate across it. There are caves, loose resources to collect, crafting, inventory, weather, a day and night cycle and the beginnings of the survival systems.

There are dinosaurs wandering around.

There is also now a proper map.

Quite a lot has happened.

The world is no longer infinite

One fairly fundamental thing has changed since I started.

Originally, Evolve had an infinite world.

Technically, it still can.

The terrain generation system is deterministic, so the world can simply continue generating as you travel. Given the same seed and coordinates, the server and every client generate exactly the same terrain.

Up until now I haven't even needed to cache generated terrain to disk. You could just keep travelling and new terrain would continue appearing ahead of you.

From a technical point of view, that works really well.

But after actually playing in it for a while, I've realised I'm not sure I actually like infinite worlds.

They sound brilliant.

In practice, I found there was something slightly strange about knowing that the world simply never ended.

There was no real sense of where I was within it.

If you travel north for a long time, there is always more north. There isn't really a shape to the world that you can learn or remember.

So I've changed it.

Evolve's worlds are now finite islands.

The generator still works in essentially the same way, and there isn't really a meaningful technical restriction on how large an island could be. If I wanted to generate something ridiculous like a thousand kilometres by a thousand kilometres, the underlying system could still do it.

But eventually you reach an ocean.

I think that gives the world something that was missing before.

It has a shape.

And because it has a shape, I can now generate a proper map of it.

You can look at the coastline, mountains, valleys and terrain and gradually build an understanding of where you actually are.

The in-game map tab showing a square island of green lowland, grey mountains, blue lakes and sandy patches, ringed by deep blue ocean, with the player marked as a yellow dot

That feels much better to me than simply travelling forever.

There is now quite a lot to do

The terrain system itself is in a reasonably good place.

There is world generation, caves, trees and foliage.

I've also added deterministic placement for smaller world items.

Things like rocks and sticks can now be scattered naturally through the world, which starts giving the player the early resources needed for the Stone Age part of the game.

The first version of crafting is also working.

I've now got a fairly substantial Tab menu as well.

Pressing Tab opens what is becoming the main interface for most of the game's larger systems.

The crafting tab with an empty inventory grid, category filters down the left and a list of hand-crafting recipes such as a campfire, cordage and a sharpened spear

There is the inventory.

There is crafting.

There is a handbook.

The handbook tab listing early tools such as a digging stick, fire striker, flint flake and hand axe, each with a short description and its recipe

There is the world map, including your position within it.

And there is an early character page.

The character page isn't doing much yet, but eventually I want it to become somewhere you can see information about your character and perhaps some statistics about what you've actually done.

Things like distance travelled are the sort of thing I'm experimenting with.

It is all still quite early, but the basic structure is there.

The weather is alive

The atmospheric side of the game is also now properly running.

There is a complete day and night cycle along with several different weather presets.

A full moon behind thin cloud above a dark stand of pine trees at night, with a flying creature silhouetted to the left

At the moment those are clear, cloudy, foggy, rain and storm.

Heavy rain falling on a grey lake with grassy terraced banks and dark hills beyond, ripples spreading across the water

Like most things in Evolve, the weather definitions are data-driven, so adding and modifying weather should eventually be relatively straightforward for both me and modders.

One thing I particularly like is the fog.

It isn't just a fixed amount of fog applied to the camera.

Fog actually moves across the world.

The fog simulation is deterministic as well, meaning the server and all connected players see the same fog moving through the same places.

You can actually watch banks of it travelling across the landscape.

Low mist through broadleaf trees at sunrise, with red-flowered bushes on the ground and stepped voxel terrain behind

It is a relatively small thing, but it has made a huge difference to the feeling of the world.

It works. But is it fun?

This is probably the more important part of the update.

The game is reaching a point where I can actually play it.

And I'm not sure it is hooking me yet.

The first day is quite enjoyable.

You arrive with nothing, start exploring, pick up resources and begin working out what you need to survive.

That part already has something.

A small campfire burning among dry scrub on a beach at dusk, with the sun low over the sea behind

The problem is what happens afterwards.

At the moment, Evolve doesn't really have a purpose.

There isn't anything pulling me into the next hour.

And that worries me much more than whether one particular technical system works.

I play quite a lot of survival games.

Valheim is probably one of my favourites. I play ARK. I play Rust.

They are all very different games, but with each of them I can normally explain why I want to keep playing.

There is always something just beyond what I'm currently doing.

A new biome.

A boss.

A better piece of equipment.

A base I want to finish.

Something another player has that I want.

Some reason to go somewhere dangerous.

Evolve doesn't really have that yet.

I can run around the world and enjoy being there, but I'm not currently thinking about it after I've stopped playing.

That is probably the next big problem I need to solve.

Dinosaurs are still difficult

I'm also still wrestling with navigation.

The voxel terrain makes AI considerably more complicated than I originally expected, particularly because the dinosaurs are all completely different sizes.

When you look at something like Minecraft or Vintage Story, a lot of the creatures roughly fit within the scale of the blocks they are navigating.

That makes an enormous amount of sense.

Evolve does not have that luxury.

A Compy is tiny.

A T-Rex is enormous.

A Brachiosaurus is ridiculous.

A brown voxel Brachiosaurus walking through a meadow of tall grass beneath an autumn tree, with dark pines on the skyline

Trying to make all three convincingly navigate the same blocky terrain is difficult.

It can also create problems for the player.

Fighting something while standing on the edge of a voxel step can sometimes feel awkward. The dinosaur may be above you, below you or trying to move onto a piece of terrain that doesn't quite suit its body size.

It works.

I'm just not convinced it feels good enough yet.

This is something I'm still trying to understand rather than something I currently have a solution for.

Steam Deck is still the target

I also still haven't tested the game properly on a Steam Deck.

I have, however, tested it on an Xbox Ally X.

On low settings I'm currently getting somewhere around 35 FPS.

That's not terrible for where the game is in development, and there are still plenty of things I know I can optimise.

It does make me slightly nervous about the Steam Deck.

If I'm getting around 35 FPS on an Ally X, I'm fairly sure dropping the same build onto a Steam Deck today would not produce a particularly enjoyable result.

Probably something closer to 20 FPS.

So I'm not there yet.

But I'm still not giving up on it.

A lot of the heavy systems can scale. Draw distances can come down, foliage can be reduced, simulation distances can change and there are still plenty of areas where I haven't really started optimising yet.

I still think I can get there.

It is just going to require some work.

What happens next

So seven months in, Evolve is in a slightly strange place.

Technically, I'm quite happy.

Most of the foundations I originally wanted now exist.

I have a world.

It has weather, caves, vegetation and dinosaurs.

I can collect things, craft things and start surviving in it.

A lit tent and campfire glowing on a dark headland by the water at night, with tall pines behind

But I'm reaching the point where the question is becoming less about whether I can build the game and more about what the game actually is.

And I'm not completely sure yet.

I have some ideas.

The next thing I'm going to do is start trying some of them.

Some will probably work.

Some probably won't.

And if they don't, I'll remove them and try something else.

Hopefully the next time I write one of these I'll have a slightly clearer idea of where Evolve is going.

For now, though, I think figuring that out is probably the most important thing I can work on.

Snow-covered voxel trees and orange autumn canopies under a grey foggy sky, with mountains fading into the distance

A lake at sunset with dark hills behind, grassy terraces on the far shore and the survival bars and hotbar along the bottom of the screen