This is the first dev diary. Posting one every two weeks from now on.

Imperium of the Stars is a 4X grand strategy game that I've been building solo. This dev diary is about the foundations: what the game is, what I'm trying to build, and why.

The game takes place in a Grimdark future, the year is 30,000. 300 years after a catastrophic isolation event called the Long Silence, the galaxy is fractured and slowly rebuilding. You pick a faction, expand across stars, manage planets and fleets, and fight your way to dominance.

Closest comparison is Stellaris. Galaxy-scale empire management, AI empires doing their own thing. But most important, I want the faction choice to actually matter. Not just different aesthetics, but different mini-games.

Pick the Imperium and you're navigating a court. Competing interests, ancient authority, an emperor who speaks rarely and finally. Pick the Void and you're managing four gods and the oblivion they feed on. Every world you touch is a resource. Pick the Grokhar and you're building up a Pack-Lord. Win battles, grow his power, hold the pack together long enough to call the Muster.

The thing I care most about: every faction should feel like a different game.

Not just different stats. Different economies, different win conditions, different ways of thinking about the galaxy.

The asymmetry goes into the UI, the resource systems, the strategic options. But the most interesting example is tech. There's no shared tech tree. Each faction has a completely different research mechanic.

Imperium has two parallel tracks. Standard research branches split into four year-gated Eras, so the tree reveals itself over 400 years instead of all at once. And a lost-tech recovery layer on top: explore relic systems, invest in a heritage building, earn Archive Fragments that unlock powerful hidden techs the empire tries to "remember."

For the Void, Favor is research. The four tech paths are the four gods, and you advance a path by doing things in that god's domain. Favor is zero-sum, the gods are jealous, so you can't please all four equally. Whisper pacts can teach locked techs at a corruption price. Hit 100 Favor and an Ascension triggers.

Grokhar is still being sketched out. Details will follow later.

All three are still in development, but the direction is the same as everything else: tech should feel as different between factions as the rest of the game.

I want thousands of star systems. Large enough that the frontier is always real. You can lose a battle on the edge of your territory and genuinely not know what's out there.

The problem is performance. This is the constraint that shapes more decisions than anything else in the codebase.

Early game is fine. Small simulation, everything runs. Late game is where it gets hard: thousands of owned planets, multiple AI factions doing their own thing, fleets spread everywhere, corruption spreading world by world, sieges ticking every day. All of that has to stay fast or the game feels laggy.

So you have to be deliberate. A naive scan of every planet every day doesn't scale. The game tracks active processes in indexed sets, so the simulation only touches what's actually changing. None of that is visible to the player.

Goal: 10k star systems, still snappy in year 400.

Two layers, both in the UI and in how you play.

Macro is the sector layer. You issue edicts across a whole sector at once: Martial Law, doubled tithes, an Inquisitorial Sweep. One decision, dozens of planets affected. For fleets and armies, macro means Task Forces. Assign fleets and armies to a task force, set a target like "invade this system" or "explore this region," and they operate. No target? They follow rules you define: invade, defend, explore. You set the doctrine, they execute.

Micro is the planet view. Build and upgrade structures, watch loyalty tick down because you overtaxed, manage corruption before it spreads. For fleets it's direct control: position ships, pick targets, manage armies on the ground.

Task Forces so you're not babysitting every fleet. Direct control still there when it matters.

I've played too many strategy games where the game was good and the experience was exhausting. Information buried in menus. Critical data not on screen. Six clicks for something that should be one.

I build for QOL from the start and keep iterating on it.

Planet filter and search. With hundreds of worlds you need to find the ones that matter fast. Filter by loyalty, corruption, resource output, whatever you need. Alerts that surface problems before they become crises, without spamming you. Tooltips that explain what a number means and where it comes from, not just label it.

The fleet invade flow is designed so you can find a target, commit a fleet, and confirm in as few steps as possible. Sounds small but it adds up.

Systems are working. Galaxy is playable. Factions are differentiated, still a lot to build.

Next entries will go deeper into specific systems.