Hi Mr Joosep Heering,
I have presented in the verbose form in the thread the base how to assemble the model already
I want to stress that the method is adaptable for simple joints as rotation one and with the limited scope of controlling all 'freedoms' of an object moving along a path.
In the case of more complex joints like a ball joint (which in fact is a gimbal joint), the situation gets a significant load of complications. The code behind is painfully strenuous (at least was for me), and even more so, it is for the digestive tract of F360 to the stomach and consequently producing the outcome blob at a reasonable pace.
The method is a part of the F360 extension I call MT123, which is not planned for public release.
As the brief conclusion,... there is undoubtedly a deficiency in the way F360 handles complex 6DoF (9+DoF) scenarios.
Nevertheless, let's hope that TF360 will wholly reveal 4D+ space beauty to its waiting diaspora ... by chance in this epoch (Enoch).
Regards
MichaelT
MichaelT