Thanks for letting me know.
@ras this is an issue with FME and we'll try and fix it. In the meantime, if you connect the output of your SurfaceFootprintReplacer to the Clipper input port then I think it will give you a workaround
@egomm Hi and thanks for the reply, I can replace the chair instances with a BoundingBox (solid), result stays the same.
I guess it's a bug, or the Clipper was never intended to be used this way?
Which is a reall bummer...
Now I wonder, if there is a way how run an inclusion test using Python. I think Scipy has a convex hull function.
But it's kind of crazy that I have to DIY. :(
It looks like the clipper geometries are actually getting rejected. I think you can only clip solids with solids. Still, it shouldn't give the appearance of having worked and features still exit via the inside port.
It seems to be a defect in the Clipper. I would recommend you to contact Safe support to clarify the issue.