The last empty corners of the map
By this point the world of The Long Watch already grew plants — but only in a handful of climates. For a long while the desert grew nothing, the cold dry grassland grew nothing, and the steamy tropical forest grew nothing at all. This is the story of the day we walked the map looking for every climate that grew nothing, and closed the last of them.
It is an unglamorous kind of work, and we mean to tell it that way. There was no clever new system at the end of it. There was a list of empty places and a quiet promise — that no living thing in this world would ever be set down somewhere it couldn’t survive — and a day spent making the second of those true everywhere the first had been false.
Why an empty corner is a quiet emergency
Our starter set of plants was tuned for the kind of country we’d been looking at most: meadows and temperate forest. That was reasonable while the world was mostly meadow. But it meant that whole climates were nothing but bare ground — a desert with not a blade in it, a cold steppe with nothing rooted, a tropical forest that was forest in name only.
On its own, an empty desert is just scenery. The trouble starts the moment something has to live there. The animals built for those places — the dry country’s grazers, the cold grassland’s hares and reindeer, the warm forest’s browsers — were arriving into land with no food at all, and so they starved almost the instant they were placed, and the hunters above them lost their prey and dwindled right behind. (That collapse of the new grazers in cold, dry, and wet country is its own story.) An empty climate isn’t a cosmetic gap. It’s a place where nothing can ever live.
Every empty corner of the map was a place we’d quietly promised something could live, and then given it nothing to live on.
The corners revealed themselves one at a time
We didn’t go hunting for these gaps so much as keep tripping over them. The way we’d been building the living world had a habit of pointing at its own next hole: you add a predator and discover it has no prey; you add the prey and discover it has no plants; you add the plants and find the next climate over is still bare. Each layer you lay down quietly exposes the one missing beneath it.
So the empty corners surfaced in order, as honest gaps rather than bugs. By the time we’d put the warm-climate and cold-climate animals in place, the message was impossible to miss: three whole climates had no producers in them whatsoever. The food web in each was a chain with nothing at the bottom. The next piece of work named itself — give every empty climate something green before anything else could stand on it.
Three recipes, all just description
So we filled the corners in three passes, and finished all three on the same day. First the warm, dry deserts and scrub. Then the cold dry grasslands. And last the tropics — the final empty climate. Five plants in each pass, a small honest cross-section of each place from the ground cover up to the loneliest tree, and the roster grew from seven species to twenty-two across the sweep.
None of it was new machinery. Each plant is written as plain description — what climate it likes, how fast it grows, how it ages and decays and casts its seed — and the world grows it from those facts alone. (That growing a plant from nothing but a handful of plain sentences costs almost no new code is its own story.) The interesting part was that each climate wanted a genuinely different recipe, even though the only thing we were writing was numbers and words.
The deserts wanted warmth and dryness, and a roster sparse enough that the dune still reads as dune — that close-up, and the five dryland plants in it, is its own field note. The cold grasslands wanted the opposite temperature but the same dryness, and a twist all their own: the ground there is so cold that dead plants come apart very slowly, so fallen wood and litter linger far longer than anywhere else. That batch leaned into a hard winter sleep and a sparse birch that genuinely sheds its leaves in autumn. The tropics wanted the opposite recipe again — plants that love warmth and wet, that mature fast, whose dead matter rots away quickly in the warm damp, and that barely notice the seasons at all, because there is no killing winter to notice: broadleaf things that stay green the year round.
A misunderstanding the tropics caught
The tropical pass corrected something we’d quietly been carrying wrong in our own plant notes. We had half-believed that the numbers describing how well a plant likes a climate were also setting how many of it would appear. They aren’t. How many plants show up is decided elsewhere — by each plant’s own density and the short list of climates it’s allowed to grow in. The climate-fit numbers do something different and quieter: they set how fast the plants that do appear grow and spread.
It’s a small distinction and it mattered a great deal. Catching it meant we went back and grounded the tropical numbers against what the system actually does, rather than what we’d assumed it did. It was the kind of correction that only surfaces when you stop and check your work against the living world instead of trusting an inherited belief — and the tropics, the last and lushest corner, was where the assumption finally broke.
Each corner, counted
We don’t trust a plant until we’ve watched it actually land. So each pass ended the same way: run a world in that climate and count what grew, where before there had been a flat zero. The cold steppe came up with around 2,712 plants on ground that had grown nothing. The tropics, the richest of the three, came up with roughly 3,640 — a lush floor for a lush place.
And the spread within each count told the truth of the place. That a plant’s role, not the climate, sets how thickly it grows — canopy always the rarest, ground cover always the densest — is its own story; here it simply showed in the numbers. In the tropics that gap is enormous — on the order of 59 giant canopy trees standing over something like 1,641 of the broad ground-cover herb beneath them, in the same patch of measured ground. A rainforest is exactly that: a scattered few giants, and a dense green floor in their shade.
The map, finally full
After the tropical batch landed, we ran the check that the whole sweep had been for — and there were no empty corners left. Every climate in the world now carried at least one living thing. The desert grows something, the cold steppe grows something, the tropics grow something; the meadow and the forest always had. For the first time, there was nowhere left on the map where life simply couldn’t begin.
One small coverage piece followed close behind, because a living world has to look after its dead as well as its living: warm-climate scavengers, so a body that falls in the desert or the tropics gets found and returned to the ground the way it already is in cooler country. The birds that finally closed that gap are their own story.
The thing we’ll remember about the day isn’t any one of the counts. It’s the shape of what the work was: not building something clever, but walking the edges of the world and refusing to leave a single corner where nothing could grow. A map can be beautiful and still be lifeless in places. Closing the last empty corners is the quiet difference between a world with pretty dead spots in it and a world that lives all the way out to its edges — every climate growing something, feeding something, letting it fall and rot back into the soil. It’s unshowy groundwork. But it’s the groundwork a game is built on when its whole promise is that you don’t win, you tend.
