Hello,
Is there any reason why ArrayCrv uses the closest point from the basepoint rather than the curve seam which would seem more consistent ?
I spent almost an hour trying to understand why copying an object from the center on a basic circle wasn’t working properly… until I moved it slightly off center (might as well move it directly on the curve in this case).
Maybe add a command option to choose the base point on curve as well ?
I agree that the Curve Seam _crvSeam is the expected start Position for the arrayCrv.
The help/documentation should maybe also mention, if the current Cplane has any influence on the orientation / rotation of the objects.
something like:
World-Coordinates (or c-plan ?) are mapped to the curve Frame…
https://docs.mcneel.com/rhino/7/help/en-us/index.htm#commands/arraycrv.htm
_divide the target curve
_orientOnCrv (copy-Yes)
(alternative use copy if no rotation is required)
make a copy of the target curve
_divide
_trim one of the segemts 
_arrayCrv on the open copy of the curve
To be honest, I would never do this in Rhino in the first place. It’s for a training exercice I’m preparing.
… well I just tried to help, based on the Info you provided.
if it is a training exercise I recommend to use a Brep, that shows its orientation more precisely then just a centred dice. I love the “Ab” textobject for those cases - as shown above - they will show as dA if mirrowed.
and will show the un-expected arrayCrv behaviour / implementation much better
kind regards - tom
Yes you are absolutely right, the cube I posted above was only a simplification to show the problem.
