By Cartesian move I mean G1/G2. I am looking for trajectory generation and comparing with ROS and KDL.and any other open source solutions.
I extracted TC/TP and removed EMC global stuff (status, configuration) and bundled into "trajectory module" using QT. I rewrote posemath functionality using Eigen (like ROS, KDL) .
Now I'd like to test against working system. Of course it didnt work perfectly (hate making all those changes at once). If there are unit tests I'd really appreciate it - didn't see and LinuxCNC usually very straightforward.
When I run my own tests there is always blending between straight line segments (see attached gif). I have tried combinations of exact stop and zero stop and no blending, but always get this parabolic blend. I am probably doing something wrong.
Need expert help.
TIA.
John
