1,510 words

The three named comparables, two more that matter, and the specific gap this project occupies.

The gap, stated first

                    REAL ASTRONOMY
                          │
        Space Engine ●    │    ● InertialRef
        (not a game)      │      (the gap)
                          │
    ──────────────────────┼──────────────────────  SEAMLESS SCALE
                          │
          Starfield ●     │    ● Star Citizen
        (loading screens) │    ● Elite Dangerous
                          │    ● No Man's Sky
                          │
                    SYNTHETIC GALAXY

Every axis of this design exists in a shipped product. The intersection does not. Space Engine has real astronomy and continuous scale and is a planetarium. Elite has continuous scale and a game and a synthetic galaxy. Star Citizen has fidelity and no galaxy. The empty quadrant is real astronomy, continuous scale, and a game — and none of the occupants can move into it without rebuilding their foundation, because a synthetic galaxy is not something you retrofit real data into.

Star Citizen — Cloud Imperium Games, 2012–, unreleased

What it is. A first-person space and ground simulation at extraordinary fidelity across a handful of hand-built star systems. Reportedly over $700M raised and thirteen-plus years in development, still without a 1.0 release [Source: CIG's published funding tracker at robertsspaceindustries.com/funding — figure moves continuously; validate before external use].

What it does better than anything. Seamless ship interiors, physicalized first-person interaction, and a sense of presence in a machine that nothing else matches. When it works, standing in a ship's hold while it maneuvers is the best thing in the genre.

What we take. Ship interiors as continuous space. First-person everything. Physical interaction with objects that have mass.

What we do differently. Galaxy scale rather than set scale. Six hulls assembled from parts rather than hundreds hand-modeled. A browser tab rather than a 120 GB install. And a defined, shippable MVP.

The lesson. Star Citizen's scope has no floor, and it is the clearest cautionary example available of what happens when a design document contains no prioritization. Every milestone in production exists partly to avoid this outcome.

Elite Dangerous — Frontier Developments, 2014

What it is. A 1:1-scale Milky Way — roughly 400 billion star systems [Source: Frontier Developments' published figure] — with the genre's best exploration loop and its best cockpit.

What it does better than anything. In-system travel pacing. Fuel-gated route planning. First-discovery credit as a reward that has motivated a decade of play with no material payoff. The discipline of never leaving the seat.

What we take. Frankly, a lot: the pip system, the A–E module grades where D is best for explorers, first-discovery attribution, the jump-range spread, and the shape of the galaxy map. These are solved problems and the correct move is to take the solutions and spend the saved effort elsewhere.

What we do differently.

	Elite	InertialRef
Galaxy	Synthetic beyond the local bubble; ~150,000 real stars in a synthetic frame	Catalog-first, procedurally filled, and it updates as astronomy does
Atmospheric worlds	Expansion content, still limited	The base case
On foot	Odyssey — built as a shooter, to a mixed reception	Built as a survey and hazard layer
Economy	Credits, grind, engineering RNG	Data only; no grind, no RNG
Install	~50 GB	A link
Progression	~8× jump-range curve, with a long cage at the bottom	~7.7×, same spread — but a steep early curve, so the cage phase is short

The lesson. Elite's most-loved system — exploration — has the least mechanical machinery in it. Its least-loved systems are the ones with the most: engineering RNG, grind gates, and the Odyssey shooter. Complexity is not what made it good.

No Man's Sky — Hello Games, 2016

What it is. A procedurally generated galaxy of roughly 18 quintillion planets [Source: Hello Games' published figure], and one of the industry's great redemption stories, having launched incomplete and improved continuously for years.

What it does better than anything. The joy of naming things. Color and strangeness. A generator whose outputs surprise its own creators.

What we take. Discovery-as-reward. Naming as the emotional payoff. The principle that generation is content rather than a substitute for it.

What we do differently. Real astronomy underneath. Real orbital mechanics and real momentum — NMS's flight is arcade by design. A genuinely continuous planet-to-space transition rather than a curated one. And a much narrower scope: no crafting, no base building, no farming.

The lesson. Two, and they point in opposite directions. Launching on a promise you cannot verify is nearly fatal — which is why the twelve capability checks are executable rather than described. And a small team sustained a galaxy-scale game for years through incremental generation work, which is direct evidence that this project's shape is viable at this project's size.

Space Engine — Vladimir Romanyuk, 2010–

The most important comparable, and the one usually left off the list.

What it is. A real-astronomy universe simulator: real catalogs, procedural fill, seamless scale from intergalactic to surface. Everything this project claims as its technical differentiator, done first and done well.

What we take. Proof that the approach works, and a great deal of prior art on catalog-plus-procedural hybridization.

What we do differently. It is a planetarium, not a game. There is no loop, no reward, no ship you operate as a machine, no reason to be anywhere in particular. It also runs as a native application and has no multiplayer.

The lesson, and it is the sharpest one here. The technical achievement this project is most proud of is already fifteen years old in another product. The differentiator is therefore not "real astronomy at continuous scale" — it is a game built on it. Any pitch that leads with the tech is a pitch that Space Engine already answers. Lead with the loop.

Starfield — Bethesda, 2023

What it is. A large first-person space RPG with over 1,000 planets, and loading screens between essentially every context — ship to space, space to surface, surface to interior.

What we take. Nothing mechanically. It is included because it is the most recent large-budget attempt and its reception is instructive.

The lesson. The most consistent criticism of Starfield was the absence of continuity — that space travel was a menu and planets were disconnected boxes. An enormous, well-funded team shipped a space game without seamlessness and the audience noticed immediately and loudly. That is strong evidence that pillar 1 is worth what it costs.

Feature matrix

	InertialRef	Star Citizen	Elite Dangerous	No Man's Sky	Space Engine	Starfield
Real star catalog	✅ and versioned	❌	Partial	❌	✅	❌
Updates with new astronomy	✅	❌	❌	❌	Manual	❌
Seamless orbit → surface	✅	✅	✅	Partial	✅	❌
Seamless ship → on foot	✅	✅	Partial	✅	❌	Partial
Galaxy scale	✅	❌	✅	✅	✅	❌
Newtonian flight	✅	Partial	✅	❌	N/A	❌
First-person only	✅	✅	✅	❌	N/A	❌
Runs in a browser	✅	❌	❌	❌	❌	❌
Playable fully offline	✅	❌	Partial	✅	✅	✅
Deterministic universe	✅	❌	✅	✅	✅	❌
Open source	✅	❌	❌	❌	❌	❌
Photorealistic fidelity	❌	✅	Partial	❌	✅	✅
Persistent multiplayer	⬜ later	✅	✅	Partial	❌	❌
Deep content breadth	❌	Partial	✅	✅	❌	✅
Shipped	⬜	❌	✅	✅	✅	✅

The bottom four rows are the honest ones. This project loses on fidelity, breadth, and having shipped. It wins on continuity, reality, reach, and openness. Every strategic decision in this bible follows from reading that table honestly.

Positioning

Elite Dangerous's exploration loop, run on real astronomy, seamless all the way down to your hands, in a browser tab.

Market gap: there is no space simulator in a browser at any level of seriousness, and there is no game — as opposed to a planetarium — built on real catalog data. The second gap is the defensible one; the first is the distribution advantage that makes anyone look.

Why now: WebGPU makes browser rendering of this kind viable for the first time; Gaia DR3 makes a real catalog of a useful size freely available for the first time; and a solo maintainer working with coding agents can produce foundational systems work at a rate that was not previously possible — which is the only reason a project of this shape is attemptable at this size.

charter — the short version
production — the plan that avoids the Star Citizen failure mode
risk — including the risk that Space Engine's existence deflates the pitch