hey fordav sunshine, are you still arround ? my pleasure to anounce that i am ready on this one, for okuma
i 1st solved this in 2019, but it was slow; better results popped up like a month ago, reaction time is pretty good
you where on the right path, and there is more to make it work
current development is like every torque function is a particular case of the method, thus instead of using a special function, you can simply code it, like spec emulation through code; g-code no longer drives, but is driven, you can't predict next movement ... and i can do other things as well
highest change rate is machine & setup specific, so such a method requires tuning the code for the machine, as it may behave different on another one, or use a slow enogh change, that is guaranteed to work on any other machine