Deep Ice — what we're building, and where it honestly stands

Two and a half thousand kilometres of ice, and a research station beneath it.

That's the setting for Deep Ice (the current working title) — a 1-to-4 player cooperative survival horror game on Jupiter's moon Europa. You and your squad are CSAR operatives — combat search and rescue — sent to find out why a research base went dark. What starts as a rescue turns into something worse: the station's crew are still there, but they aren't themselves any more.

That's the game. Here's the honest part.

One person, building in the open

Night Wraith Studio is one person. I'm a software engineer of twenty years — but until this year, never a games programmer. The C++ and the Unreal Engine behind Deep Ice were both learned from scratch to build it. What wasn't learned from scratch is the engineering discipline underneath: the architecture, the networking, the automated tests, and the self-hosted build pipeline are two decades of systems work pointed at a new medium.

I'm building this in the open, and this blog is where that happens — the design decisions, the engineering, the things that break, and the occasional thing that works first time.

What's real today

I'd rather show you the honest state than a trailer full of things that don't exist yet. Right now, Deep Ice is a working game in greybox. Two players on separate machines can drop into a session over the internet — no servers I run or pay for — form a squad, travel into a mission together, fight through it, feel the station's lethal cold close in, and talk to each other by proximity voice as they go.

The systems work. The art doesn't exist yet: everything you'd see is grey boxes and placeholder UI, on purpose. Systems first, paint later.

The thing I most want to get right

Plenty of co-op games boast about "adaptive AI." Play most of them five times and you'll spot the pattern — the spawns are a shuffle, not a response.

The heart of Deep Ice is a director that actually reads how you're playing: whether the squad bunches up or spreads out, moves efficiently or struggles, plays it safe or pushes too hard — and paces the horror around that. A run where you stick together should feel nothing like one where you scatter. No two will ever play the same.

It's the part I'm most set on getting right, and you'll see a lot of it here as it comes together.

Follow along

I'll be posting as the work happens — deep-dives on the engineering and the design, and the honest ups and downs of building a horror game solo. If that sounds like your kind of thing, subscribe below.

Something's wrong under the ice. Come find out what.

— Chris, Night Wraith Studio