a thing i often find myself doing in factorio is building a new coal mine to fuel my smokestacks to power my ammunition fabricators to kill all the animals on the planet.

i also often find myself thinking "this is fucking awful"

finally, i then wonder how i could make a morally redeemable version of the game based on ecosystems, until i get stuck at the part where the point of an ecosystem is that it doesn't require effortful intervention, and go back to throwing grenades into nurseries.

terra nil is South African studio Free Lives' attempt to get past the paradox.

in factorio, you are stranded on a planet and must construct from scratch the technological capacity of a modern nation state to build a rocket and get off the planet. how this plays out is: you're a little guy running atop patches of ore, connecting densely packed machines with pipes and conveyor belts. due to a judiciously designed material versimilitude — everything you build requires its own raw materials and energy inputs — the very act of pursuing your goal more efficiently necessarily forces you to balance the consumption and production of your factory.

this improved conveyor belt requires more steel. i'll need to increase my production of iron and reroute my coal supply blah blah blah...

ultimately, you kill most of the animals that were there first and industrialize enough of the planet to launch off into space again. as negatively as i'm framing it here, i will say that it is very, very satisfying to do this.

in terra nil, your objective is nominally the same: work on a planet and leave it. the difference is that balancing resources is the explicit goal. the planet you land on has already been desolated, and it's your job to restore a healthy ecology. you do this by placing buildings on poisoned soil that work synergistically with each other to clean, fertilize, and germinate the land. doing this optimially requires you to tetris your placements correctly in order to minimize redundant overlap, prioritised in different ways depending on which stage of the level you're on - detoxicification, biome restoration, and fauna reintroduction. once you've done these three things, you also have to recycle every building you placed, leaving only footprints as you fly off.

some more thoughts on the differences between the games

• in factorio, alongside what's there to mine on the planet, your own time is a resource because everything has to be done or set up manually. it is amazing how close it can get to the "this is work" boundary without making you want to stop playing, because you can't accomplish your ultimate goal without sinking a lot of time into the game. in terra nil, there's a currency system and buildings appear immediately as you click on where you want them to be. thinking it through takes time, too, but far less time is spent on execution of your plan, comparatively. i think there's a lot to be written here on how controlling a guy versus birdseye strategy game agency affects the work/play boundary but if you're interested in that you should just go read an actual game theorist.
• factorio's planets aren't beautiful. despite abundant water and minerals, most of the planet is brown. the animals there are basically the zerg, with no apparent culture or behaviour beyond "swarm and attack". you could take pride in this - "graphics aren't the point. it's about the gameplay" - but i think this should be understood as more than a simple consequence of game development prioritisation. the planet isn't beautiful because makes it less disturbing to destroy. factorio players grimly joke about how the factory must expand, because it's already uncomfortable enough. imagine if the land was diverse, lush, and thriving. or if the animals all had names and cultures and complex societies. it would be awful. it would be like a first-person shooter where all the enemies are children. so obviously the designers haven't done that. but in terra nil, beauty coming back to the land is probably the largest positive reward it gives you (i don't really care about getting a percentage bar to 100%). aesthetics are quite inseperable from gameplay.

so, i think terra nil's shift to something more akin to a puzzle, with a focus on aesthetic objectives is an interesting way to handle the paradox. it is a far shallower game. it doesn't teach you much in terms of fundamental principles of the universe in the way factorio does, but it was a calming and fun way to spend an afternoon and i think the designers of the future eco-factorio could take a lot of good lessons from its successes.

(i was clued into terra nil via the honourable mentions appended to Mark Brown's analysis of Jusant as this year's most innovative game. Thanks, Mark!)