I’ve been thinking about a traffic system in theory. The idea I have sounds good to me but there might be “hidden traps”. I would be glad for your opinion/suggestions before I start implementing it as a whole. I’m mostly worried about the performance issues.
So, the path system would consist of many splines (image) and when finishing a spline, the car would switch to one of the appropriate splines.
The splines themselves would be stupid, they would only exist, contain data of their appropriate splines (to give the car a choice where to continue) and maybe two booleans for controlling turn signals.
The car would maintain a reference to the current spline and it would operate the spline’s "location closest to car’s forward vector). It would navigate using “delta of look at spline and car’s rotation”.
It would sense other cars using line trace, with length dependent on the speed. If I get cocky, they would overtake slow ones if the sensing was good enough.
The traffic wouldn’t be too high.
I’m mostly concerned about the number of splines and the solution of following the spline. (Maybe I’m just overly worried about the performance, I’m still used to much weaker engines.)
I admit, I haven’t seen any of the traffic tutorials. Since I already have a solution in my head, I’m quite unable to pay enough attention to those tens of minutes long videos.
Thank you in advance for any replies.