On a GeomTransform:
getPostition(float[] dst) gives the “parent Bodys” postition
getPosition(Vector3f dst) gives the transformed position.
Is this “by design - not just yet documented” or a bug? According to ODE docs, dGeomGetPosition returns the bodys postition. Which one is the most “future-proof”?
/M
ah.
I have found GeomTransform’s most confusing in the past. The documentation is pretty scarse on them, and you have to be careful with the order of your calls or else it may not work.
The GeomTransform object isn’t 1:1 with it’s ODE calls. The reason I made it give the transformed position is so that bound 3D geometry is drawn in the correct location.
These methods are subject to change. You are more than welcome to suggest how we should fix this. Regarding “future proofing”, don’t worry – both capabilities are needed and won’t be removed. The method signatures may change, but you should be warned if this happens.
In short, this is a combination of “by design”, “not yet well documented” and “a bug”.
I’ll add this to my list (it is a long list…).
Will.
It’s kind a hard in the beginning (first 50 years?) to know where the stuff ends up in ODE, if you put the things together correctly etc etc, so I just started on a simple class the takes an ODE World and renders all bodies in it (simplified, with no optimizations what so ever, using LWJGL atm). Should I try to make those 50 lines of code public? (Or, perhaps more importantly - is there already code that does this?) At least, I would have spared a few hours getting that code from the community…
But the issue at hand - having the opportunity to get both the “world position” (transformed) and the parents postition/relative postion/… feels nice, all needed is just a clarification of the API - like getWorldPosition will always give you world position - no fuss, while getPosition will give whatever the ODE docs states!?!