Inner workings of Network Ownership

I am attempting to create a server authoritative prop system. I want to make it immersive in the way that using network ownership would be, so no delays. But I also want to prevent exploiters.

For making this system I want to start simple and just recreate network ownership to get the general idea of what I’m working with and edit it from there. So Im wondering how exactly does Network Ownership work? Would sending a CFrame to the server every frame then making the server replicate that CFrame be the exact same as network ownership both functionally and more importantly in terms of speed?

1 Like

Network ownership is for physics based things. Physics has special properties on the engine side, for example there is built in interpolation (so it’s not choppy) and it will be able to interact with physically simulated stuff properly (ex: collisions should impart momentum). So TL;DR: sending a CFrame would work but it wouldn’t be exactly functionallity identical without your own interpolation system (or perhaps there is a way to “set” the cframe of a physics part, like through an align position/orientation) and a way for it to work with the rest of the engine physics.

Do you mean by having client ownership or having server ownership? I would just use the built in ownership, unless you’re doing something special like deterministic replication (replication with/through prediction). What exactly is the use case you have that might need a custom system?

You can get the same amount of speed using unreliable remote events.

Edit: Yeah, you need unreliable remote events for the same speed:

Physics. Physics networking is unreliable and unordered, and it is replicated at 20hz. An important note is that physics is interpolated at 60hz. This means that it, under normal circumstances, has less latency than RemoteEvents


Thanks, the provided sources will be very useful.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.