Closed TheDuckCow closed 1 year ago
FYI @bdog2112 I've assigned this to you since this is the highest near-term value add change to improve. If it looks too lofty to work on, you could still work on one of the other lower hanging fruit bugs instead. Either way, happy to triage tomorrow and peer program a bit if you like.
Edit: If you've not read this page yet already, I do highly recommend it.
Appears to be the issue of using a distance magnitude instead of radians (???) for the curve tilt assignment, should actually calc the radians about the z-basis axis of the roadpoint wrt the given end curve3d point.
Currently, we have smooth interpolation from one RoadPoint to the next in terms of some lane/overall width. However, rotation about the Z axis of RoadPoints is not smoothly interpolating across a RoadSegment, per image below:
A bonus change would be to have the rotation be somehow influenced by the length of the gizmo handle (so if the handle is shorter, it will get to the 'middle point' rotation level sooner, being the halfway point between the rotation values at both ends of the road segment.