Quote:unwind the rope for a great length and go around many obstacles with it, and then unwind it, pulling the object at the end of the rope to its beginning.
Sounds to me like you need the ninja rope from the "Worms" series, the fishing rod from "Umihara kawase", or the rope from "ZenBound".
https://www.raywenderlich.com/348-make-a...ity-part-1
https://www.youtube.com/watch?v=FDt4ulUnBrc
https://www.youtube.com/watch?v=vPEvAxjay1s
They all have in common that
they're not actual simulations, just a situational trick: Only the last segment of the rope is simulated (the one that goes from the first contact against the world to the player). The rest of the rope is just a line. As the rope wraps around objects, the rendered line is lengthened and the first contact point updated. This way the "rope" can be arbitrarily long, as only the last segment is ever processed.
This does have heavy limitations though: the rope must *always* be 100% tense, no friction, no self-collisions, no damping, no two-way coupling with other objects, etc. This is fundamentally different from what Obi does.
Quote:- how long can a rope be made without losing productivity?
Depends on how many particles per unit length you need (resolution), how small a timestep length you can afford, and a few other variables. There's not a fixed maximum length, however it is often impractical to have a rope more than a few (4-5) meters long.
Quote:- what are the benchmarks (from the performance evaluations on the site I could not understand this), how to pull, how long the rope in a meter of Unity can be created?
The benchmarks in the web are
performance benchmarks, done by progressively increasing the amount of particles while keeping the timestep length. When having larger ropes, performance is typically not an issue, but convergence (simulation quality) is. Longer ropes tend to be more rubbery/stretchy. To fix this, you need to chop time in finer slices (use a smaller timestep) but this makes the simulation more costly.
If you have a low-resolution rope, say 3 particles per meter, you can easily get 40 meters long rope or longer. If you have a thin, high-resolution rope (+50 particles per meter) and must use a large timestep value, then you might not be able to go past 2 meters without the rope becoming too rubbery.
Note this is not specific to Obi:
all physics engines have the exact same limitations as any computer simulation must be discretized both in
space (resolution) and
time (timestep size). As you throw more bodies into the simulation and chop time in finer slices (increase resolution and decrease timestep size), the computational cost increases. It's a basic consequence of how digital computers work.
Code:
- if there are video clips with tests of the maximum length, I will also be grateful.
These use rather long ropes:
https://www.youtube.com/watch?v=7ugSN7laNd4
https://www.youtube.com/watch?v=Czy6WnIWtyY
I don't know if they will be meaningful in the context of your game, though.