(2026-05-09, 6:37)Peter Blomberg Wrote: The rotation matrix is [-6 0 0; 0 -2 0; 0 0 -2], which is still fully within the mirroring, scaling, and 90-degree rotation realm. This is not a box2-5 rotated to 45 degrees. Therefore, I do not see a need for a new primitive in this case.
However, it seems the other tri3 tri4 etc are also in the same non-45 degree rotation state, so my arguments are moot.
Well the new primitive adds two additional surfaces that can be eliminated from the part. This just happens to be one part where such a primitive could be useful.