Nether Earth: The RTS Before Anybody Named the Genre

Nether Earth came out in 1987 for the C64, Amstrad CPC and ZX Spectrum, developed by Icon Design Ltd and published by Argus Press Software. Nobody called it a real-time strategy game, because the term didn’t exist yet in any form the industry had agreed on. That vocabulary gap is a large part of why the game gets left out of most histories of the genre, which tend to start the clock with Dune II in 1992 or, at their most generous, Herzog Zwei in 1989. Nether Earth was doing the load-bearing parts of the same idea five years before Dune II and two years before Herzog Zwei, on hardware with a fraction of either game’s processing headroom.

Build a robot, lose the factory

The premise is a war fought entirely through robots, waged against a race the game calls the Insignians. Factories scattered across the isometric battlefield — some already yours, some the enemy’s, the rest neutral and open to capture — manufacture the two halves of every unit: a locomotion type (legged, tracked or anti-gravity, each with its own trade-off between speed and stability) and a weapon (cannon, missiles or phasers, each suited to a different range and target). There’s no tech tree gating any of these combinations; every option is available from the first factory you hold, which means the entire strategic layer is about matching a build to the battlefield situation in front of you rather than unlocking access to a build over time. Different factories on the map are specialised toward different resources — electronics, chassis parts, one weapon type or another — so which factories you hold determines which builds you can actually produce, not just how many.

Losing a factory to enemy capture doesn’t just cost you production capacity. Losing all of them ends the game outright, which turns every neutral factory into a contested resource worth fighting over from the opening moves, and turns your own factories into something you have to actively garrison rather than passively trust to survive. This is the mechanic doing the real strategic work: a factory left undefended for even a short stretch is a factory you can lose permanently, and in a war fought over a small number of production sites, a single permanent loss can be close to fatal.

Why the capture loop still works

What makes Nether Earth read as a genuine ancestor of the genre rather than a curiosity is how completely the design commits to simultaneous, unpaused decision-making. You’re managing production, unit composition and battlefield positioning all at once, in real time, with no turn structure to fall back on for breathing room. That’s the actual proposition later RTS games would formalise — Dune II: the RTS template gets credited with inventing the base-building-plus-unit-command combination because it packaged the idea cleanly enough for a mass audience to recognise it as a new thing. Nether Earth had already built the combination; it just built it on hardware too limited and for an audience too small to make the label stick outside Britain.

The capture-not-destroy objective is the sharper idea, and it’s one plenty of later strategy games never fully commit to. Most RTS design treats an enemy structure as something you demolish. Nether Earth treats a factory as something you want intact and working for you, which changes the tactical calculus completely — you can’t simply pour fire onto an enemy factory until it’s gone, because gone isn’t the goal. You have to close in, hold the position, and take it, which means your attacking force has to survive contact rather than win an exchange purely from range. It’s a harder design problem than demolition, and Nether Earth solves it convincingly on a machine with a few kilobytes to spare.

The interface is the real feat

None of this would matter if the game were unplayable, and the interface is where the design earns real credit. Rather than a mouse-driven cursor selecting units from a fixed overhead viewpoint — a solution that wouldn’t be practical on this hardware in 1987 anyway — the player controls a flying craft that has to travel to each robot in person to issue it an order. That single choice does a lot of quiet design work: it means command isn’t free, attention has a physical cost measured in travel time across the map, and a robot given an order will keep carrying it out — advancing, attacking, holding — until its objective is won or lost, because there’s no instant recall available while your command craft is busy somewhere else. Directing several units well means planning a route for your own attention as carefully as you plan a route for your robots, which is a genuinely different kind of strategic problem than clicking a unit and issuing an instant order.

The comparison worth making

Supremacy: the empire game with one bad decision in it arrived three years later and buried its real challenge in an economic feedback loop with a long delay between cause and effect. Nether Earth sits at the opposite end of the pacing spectrum — every decision resolves close to immediately, a robot either takes the ground or it doesn’t, and the tension comes from acting under time pressure rather than from a debt you don’t notice accruing. Both games are honestly about resource allocation under constraint, and both predate the genre conventions that would later formalise their ideas into recognisable templates. Seeing them side by side says something about how much genuine strategy-design thinking was happening on British home computers years before the American PC market got credit for inventing the vocabulary.

Herzog Zwei, released on the Mega Drive in 1989, is the game usually cited as the real proto-RTS, largely because Sega’s marketing reach and the format’s larger install base gave it a visibility Nether Earth never had. The Sega game adds a persuasive twist of its own — a single controllable unit that also functions as the player’s mobile command point — but the core proposition, real-time production and capture under simultaneous threat, is a proposition Nether Earth had already shipped two years earlier on considerably less capable hardware.

A design nobody had a name for yet

Part of why Nether Earth vanished from the genre’s official history is that Icon Design wasn’t working from a template — there wasn’t one to work from. Strategy games on eight-bit home computers in the mid-eighties were mostly turn-based, largely because turn structure is the easier problem to solve on limited hardware: you don’t have to worry about the machine keeping pace with a player who wants to issue several orders in the same second the enemy is issuing several of their own. Building a simultaneous, unpaused system on a C64 in 1987 meant solving problems around input handling, unit behaviour and battlefield state that the rest of the strategy genre wouldn’t formalise proper solutions for until dedicated development budgets caught up years later. That the game holds together as well as it does, on machines measured in kilobytes rather than megabytes, comes down to a tightly scoped design — a handful of factories, a modest roster of locomotion-and-weapon combinations, one battlefield at a time. That narrow scope is what made the ambitious real-time part of the equation actually tractable.

The visual presentation reinforces the sense of a battlefield rather than a menu screen. Robots move across the isometric terrain at a pace that reads as deliberate rather than sluggish, and watching a robot roll out of a factory already assembled from its chosen parts does real communicative work, showing you exactly what you built rather than asking you to read a stat block and trust it. That small piece of feedback matters more than it sounds: in a genre that would spend the next decade arguing about how much information to surface to the player and how, Nether Earth solved a slice of that problem simply by making a robot’s composition legible at a glance.

Why it stayed a footnote

Distribution is probably the larger reason the game never built the reputation its design deserves. Argus Press Software was a comparatively minor publisher next to the labels that would come to dominate British software retrospectives, and a genre-defining idea released quietly on a mid-tier publisher’s slate, without the marketing push a Sega first-party Mega Drive release could command two years later, simply doesn’t get remembered the same way, however sound the underlying design turns out to be on replay. It did win Your Sinclair’s Megagame award in the month of its release, which suggests contemporary British players recognised something unusual in it even without a name for the genre it was inventing. Nether Earth’s absence from most potted histories of the RTS says more about who got to write those histories, and which publishers and formats they were paying attention to at the time, than it says about what the game actually accomplished.

What actually plays now

Emulated versions run cleanly on any of the original eight-bit platforms, and the game is short enough in scope — one map, a small cluster of factories to fight over — that a single session gives you the whole design rather than a slice of a much larger campaign. An open-source remake exists for modern desktop systems for players who’d rather skip the emulation step, and it preserves the original’s rules faithfully enough that the lesson about attention-as-a-resource still lands. The command-craft interface takes longer to get comfortable with than a modern strategy game would ask of you, and it’s worth persisting through that discomfort rather than dismissing it as dated; the friction is doing real work, forcing the kind of prioritisation under pressure that a slicker interface would let you avoid.

Icon Design didn’t build a lasting studio identity around the idea the way some of their contemporaries did with other genres, and Nether Earth never got the sequel or the wave of imitators that a genre-founding release usually attracts within a couple of years. It sat on the shelf, a single well-made statement of an idea, while the wider industry took most of a decade to arrive at the same proposition again through different routes and with far larger budgets behind it. That’s a common shape for a genuinely ahead-of-its-time design: the credit tends to accrue to whoever ships the idea with the marketing reach to make it stick, not necessarily to whoever solved the hard problem first.

Spoilers below

The strongest opening move available is rushing the nearest neutral factory before the enemy commits resources to it, because early production advantage compounds faster than any single robot build ever will — an extra captured factory lets you run more than one production line at once while the opponent is still deciding how to split defence between their existing base and their own expansion. Sitting back and building a purely defensive force at your starting factory instead, which feels like the safer opening, tends to lose to any opponent making that early rush, because by the time you’re ready to expand outward, the map’s neutral factories are already gone and your locomotion-and-weapon options have narrowed to whatever your one remaining factory happens to produce.