Update — September 14, 2026: The architecture behind our first commitment has changed since this article was originally published. We no longer describe Stellar Mariner as relying on seamless server handoffs or dynamic server meshing to keep everybody in one world. The commitment itself has not changed: there is one persistent world underneath the simulations used to run it. I have updated the explanation below to reflect that distinction rather than leave an architectural assumption standing as though it were part of the promise.
We published a commitments page before there is a finished Stellar Mariner, before there is a store, and before anybody has paid Rynua for anything. That may seem early. To me, that is exactly why it matters.
Promises are cheap when keeping them costs nothing. They become more interesting later, when a decision would be easier, faster or more profitable if the promise were quietly interpreted a little differently. I would rather remove some of that room now. These commitments are public, dated and specific enough that somebody can quote them back at us years from now if what we build starts drifting away from what we said.
I am writing this article in my own name because that responsibility ultimately sits with me. Rynua can change technology, revise designs and discover that an assumption was wrong. My job as founder and creative director is to make sure those changes do not quietly take the project somewhere that violates the principles we said would remain fixed.
One persistent world
The first commitment is that Stellar Mariner has one underlying world. That does not mean every player must always occupy the same simulation instance, or that one enormous server process has to simulate everybody at once. Those are infrastructure questions, and our thinking on them has already evolved since we first published this article.
What matters is the persistent state underneath those simulations. Ownership should not fork because two groups were placed in different instances. A structure, claim or durable object should refer back to one authoritative state. The economy should not become a different economy because somebody happens to be hosted in another region. Simulation can be instanced, geographically distributed or changed as the architecture matures; the persistent world should remain singular.
That is a better definition of the commitment than the one I originally wrote. The first version talked too confidently about seamless server handoffs and machines moving players between one another before they met. Those were architectural assumptions, not principles. The principle is simpler and stronger: we instance the people when necessary, not the world they belong to.
A history that survives
Persistence means more than saving your character between logins. Stellar Mariner is intended to carry forward the parts of its history that give what players do meaning: ownership, structures, discoveries, reputation, economic activity, losses and other lasting consequences of the world changing.
That does not mean every piece of simulation state is permanent. A projectile does not need to exist in a database forever because it was fired once, and temporary effects do not become valuable simply because they can technically be stored. The important distinction is between what the live simulation needs now and what the universe still needs to know afterwards.
There is also an important pre-launch exception. While Stellar Mariner is being tested, there may be wipes between builds or test phases because foundational systems are still changing and sometimes need a clean state. We will make those phases clear when they happen. The commitment begins once the persistent universe is live: from that point, history is supposed to accumulate rather than being periodically cleared to reset the playing field.
Stories created by systems and players
I do not want Stellar Mariner to depend entirely on somebody at Rynua deciding what should happen to you tonight. Authored stories, locations and characters will have a place in the game, but the universe should also be able to produce situations because its own systems created the conditions for them.
A shortage can create demand. Demand can make a route worth flying. A valuable route can attract traders, escorts or pirates. A loss can create work for somebody who was nowhere near the event that caused it. The more of those relationships the world can sustain, the less every interesting evening has to begin with a designer placing a mission marker in front of somebody.
That is what I mean by stories created by systems and players. It is not a promise that everything will be procedural or that authored content disappears. It is a commitment that the state of the universe should matter enough to create opportunities and consequences of its own.
No pay-to-win
Stellar Mariner will need to fund itself. If it reaches the point where Rynua can become my full-time work and eventually support the people needed to build it properly, that money has to come from somewhere. I have no interest in pretending otherwise.
What I do want to rule out is selling competitive advantage as a business model. Spending more money should not make somebody fundamentally better at the game than somebody who earned the same capabilities by playing. A larger ship can be more capable in some situations, but size should come with costs: crew, logistics, operating requirements, handling, exposure and whatever else makes that ship a different choice rather than a straight upgrade.
The same principle applies outside individual items. We should never make the normal game deliberately unpleasant so that paying removes the inconvenience. If a mechanic becomes tolerable only after somebody spends money, then the problem was designed into the mechanic. That is not a line I want Rynua to cross.
Nothing sold on the strength of a render alone
This commitment changed substantially while we worked on it, because the more rigid version sounded responsible without being something Rynua had yet earned the right to promise.
I originally wrote about fixed delivery windows based on ship size, mandatory testing stages and automatic store rules if those windows were missed. The problem is that we had not yet demonstrated a production pipeline capable of supporting those guarantees. Publishing precise limits before proving the process behind them would have been another kind of speculation.
The current commitment is more fundamental. If Rynua ever takes money for a functional item before that item is finished, the item must already exist as something real in our build. A concept image is not enough. A specification attached to an idea is not enough. A placeholder that has never functioned in the game is not enough.
Before somebody pays, they should know what the item is, what it is for, how far through development it genuinely is and when we expect it to arrive. Its progress should remain visible after the sale. If we later establish reliable delivery windows, those windows should come from a production pipeline we have actually demonstrated rather than one we predicted years beforehand.
This commitment exists because I know what the other side of that transaction feels like. I do not want Rynua funding Stellar Mariner by asking people to buy increasingly convincing representations of things nobody has built yet.
You get what you bought
The commitment beside it is simpler: what we sell should arrive as what we described.
Games change, and balance changes constantly. Damage values move, handling changes, capacities are tuned and systems evolve as the game develops. None of that means an item can quietly become something fundamentally different from what somebody paid for.
If Rynua sells something as a particular class of ship with a particular purpose, physical scale and set of defining characteristics, those things matter. If a change becomes necessary that alters what the item actually is, the buyer should be told rather than discovering it at delivery. Whatever the legal minimum requires, I want the principle stated plainly: redefining something between sale and delivery is not acceptable simply because doing so became convenient.
This is not about freezing game design forever. It is about distinguishing balance from identity.
Why put any of this in writing now
All six commitments come back to the same idea: fair play.
Fair play between players, because money should not decide who gets an artificial advantage. Fair play between Rynua and the people supporting it, because we should not sell something on one description and deliver another. Fair play with people's time, because the universe should remember enough of what they did for that time to have consequences. And fair play in how we talk about development, because a design should not be presented as a proven system simply because the distinction is inconvenient.
None of these commitments proves anything today. They will only become meaningful when keeping one of them costs us something. A deadline may become uncomfortable. A monetisation option may look attractive. A technical shortcut may make one persistent world harder to preserve. That is when the page matters.
I also expect the implementation underneath these commitments to change. It already has. The one-world architecture described on September 14 is not the architecture I described twelve days earlier, and that is exactly the kind of distinction I want this development record to preserve. Changing our mind about how to keep a promise is not the same thing as changing the promise.
If one of the six ever genuinely needs to change, I do not want it edited quietly and left for somebody else to notice. We should say what changed, why it changed and what that means for the people who trusted the earlier version.
My name is attached to that standard.
I am comfortable being held to it.
