Ringworlds and orbitals are significantly bigger and harder to build than O’Neill cylinders. O’Neill cylinders are potentially feasible within a century or two; ringworlds and orbitals would take much longer.
>O’Neill cylinders are potentially feasible within a century or two
Is there anything fundamentally holding us back from building one right now? I mean there's the obvious problem that nobody has a good reason to build one, and only a hand full of entities have the funding to do so. But the engineering challenges seem to be on the level of building LHC or ITER: not nessesarily easy, but very doable.
SpaceX is claiming in the neighborhood of $1500/kg to put something into LEO. If you want to put a structure on the order of billions of kilograms into orbit it would make more sense to get that mass from something already out there and park it in high orbit instead - and space-based mining technology doesn't exist yet.
At the same time a fully loaded Falcon 9 uses about $10/kg of fuel to put something into LEO. The cost mainly comes from building the rockets, and to a lesser degree operating them (mission control, landing ships, etc). All of those costs would come down dramatically once you reach the scales nessesary for building such a structure. Economies of scale could easily bring the price down an order of magnitude or more.
For a small-ish cylinder that might be enough. For a cylinder that's multiple kilometers long we would want to do asteroid mining and rockets engineered to bring stuff up cheaply (at the expense of loosing a few rockets, loosing dirt is a lot more acceptable than loosing billion dollar satellites).
I wonder if we could circularize the orbit of a near earth asteroid like https://en.wikipedia.org/wiki/3908_Nyx which is outside of earth orbit always (less likely to run into the earth, and 10% farther up the sun's gravity well.
And something like $300 per kg for Starship (on the assumption of roughly $30M for a fully reusable launch - Musk says it should be cheaper than F9 per launch - and 100 tonnes to LEO) if they're successful in minimizing refurbishment between launches, and there's enough business to spread the development costs over enough launches.
They're enormous, and enormous things are difficult and expensive to build. They're also in space, and building things in space is difficult and expensive. So that's quadratically difficult and expensive.
