The Audience Was Already Built
Every new platform has the same first problem, and it isn't technical. It's that nobody makes anything for an audience of zero. A creator weighing where to spend a weekend will pick the format with an audience already in it, every time, because effort spent on a platform nobody uses is effort that evaporates. So the platform launches with no content, which means no users, which means still no content. The failure isn't a bug in the product. It's the room being empty and staying empty because empty rooms don't attract people who fill rooms.
The usual fix is to throw money at the gap: launch grants, exclusivity deals, a contest with a cash prize, a studio paid to backfill the catalog before anyone shows up. This works, sort of, the way a transfusion works. It buys time, not blood supply. The moment the subsidy stops, you're back to the same math: a platform with no organic reason for anyone to build for it, now also broke.
There's a cheaper fix, and it requires giving something up first: stop building a platform at all. Build a thing that implements a standard, and pick a standard that already has decades of homebrew sitting behind it, made by people with no commercial stake and no expectation of an audience, who built it anyway because the format was fun to work in. That population solved your cold-start problem years ago, for reasons that had nothing to do with you.
CHIP-8 is the clean case. It's a 1970s interpreted instruction set, originally meant to make simple games easy to write on machines with almost nothing to them. It has no company behind it, no licensing terms, no owner to negotiate with. What it has is fifty years of hobbyists writing small games for it purely because the spec is tiny and the result is satisfying, a body of homebrew numbering in the hundreds of programs, all free, all compatible with anything that correctly emulates the instruction set. Build a piece of hardware that runs CHIP-8 and you inherit a full library on day one. Build a piece of hardware with your own bespoke API instead, and you get the empty room, because why would anyone write a game for a console that has sold to nobody yet.
The asymmetry is almost embarrassing once it's stated plainly. A platform you invent has to earn every piece of content from scratch, one creator at a time, against the gravity of every established place that creator could spend their time instead. A platform that implements an existing standard inherits content that was made under a completely different incentive, often made by people who never considered that their format would get a new runtime decades later. You're not recruiting an audience. You're discovering one that was built without you and has been sitting there the whole time, unclaimed by anyone with a product to sell.
This only works on the standards that happen to have survived as homebrew, and that's a narrow set on purpose. It skews toward formats that are small enough for a hobbyist to implement alone, open enough that nobody had to ask permission to build for it, and old enough to have accumulated a real body of work instead of a scattering of weekend projects. You don't get to pick a trendy new format and expect the same effect, because trendy formats haven't had time to collect decades of people building for free. The option is rare, which is exactly why most platform builders skip past it without noticing it was there.
The part that actually stops people from taking this path isn't a lack of candidates, it's what the choice costs them. Implementing someone else's standard means the catalog isn't yours, the API isn't yours, the credit for the format's existence belongs to a scene you didn't start and can't claim. You get a working platform on day one, at the price of never being the one who invented it. Most people building a new platform want the second thing more than the first, which is probably the real reason the empty room keeps staying empty.
The zero-to-one problem has usually already been solved, by someone who wasn't trying to solve it for you.