Quote Originally Posted by savage
I have not used GLScene so I'm not sure that would be the best way to sandbox it as GLScene may be doing things *its own way* that may not be directly compatible with Newton, hence the need for the OxNewteon component.

I could not find what was the problem. If I were to look into it further I would send an email to Sascha as I'm sure he would be more than happy to help with the intricasies of Newton.

Also take a look at http://newton.delphigl.de/ and in particular his http://www.delphigl.de/files/glNewtonCarDemo2.zip . Maybe that might shed some light. It's straight OpenGL ( IIRC ).
i don't think i explained what this is properly, so i will try to again:

this is not a normal newton vehicle joint vehicle simulation, NGD vehicle joint uses rigid bodies for wheels, and therefore a such simulation has its drawback: instability and high processing need.

This project is implementing a raycast car for use with newton, raycast / hovercraft car instead of using bodies to simulate wheels, use raycasting where the wheel should be, and simulates the tire reactions, this is much less intensive way and more stable to simulate vehicles.

The code is implementation of RCV / raycast vehicle in C, i am trying to translate it to pascal, C version uses some utilities from newton sdk's own vector library which are probably not working in my translation. in delphi translated implementation, glscene is only used to render the vehicle, i don't use glscene's vector libs or anything else from glscene in the vehicle sim. code.