Slicing the model branches_16_closed_mid_res results in paths that are not always closed about compas_slicer HOT 1 CLOSED

Ok, I found the problem. B
oth slicers were identifying all paths as closed, so for all paths it was is_closed = True. (For cgal slicer all paths are closed by default, for default slicer it checks the mesh connectivity).

The problem was instead in the order of post-processing operations in the Path + Base_slicer class. First we closed the path by duplicating first pt, then we changed the seam position and orientation. This is a bad idea, because like that we can end up with a duplicate pt in the middle of the path, and other than that the path seems to be open, although it has is_closed = True, and it has been closed. So I believe this could explain both the holes and the mysterious duplicate pts warnings that we have been having.

I will make a PR to master with the fix. Basically I move the closing of the path, from the Path class to the Base_slicer post-processing functions, so that it happens after everything else has happened. This seems to me to solve the problem.

Let me know if u see sth similar again. nice catch!

from compas_slicer.