QuIK: A ultra-fast and robust generalized inverse kinematics solver for C++/ROS2

no need to apologise for anything. I was just curious about the difference in the nrs quoted.

I believe the biggest advantage optimisation-bases solvers have is they can work with ‘any’ kinematic configuration (within reason) – not just the one they were written/generated for.

The paper already mentions this and mentions the main case where this is advantageous: calibrated mechanisms.

Again, great work and thanks for making this available.