# MESH MYSTERY - Make Face Vertex Order Consistent

Can anyone explain why the face boundary offsets on this mesh are inside on most of the mesh faces but outside on the triangular mesh faces with naked edges?

I’ve tried unifying mesh normals and face windings and tried K2’s Mesh Direction but it gave an invalid mesh.

Are there any tried and tested ways of ensuring a mesh has all it’s face vertices listed in the same direction around the face?

This is a bad mesh in that it has 2 quad faces and the rest are triangular.
MESH_MYSTERY.gh (13.5 KB)

There is no “inside” or “outside” of a 3d curve, even if planar.
Offset function can use an optional plane to determine the direction of the offset. Use Planar to get a plane for each triangle and use that.

Or… you can use Offset Loose and bypass this problem.

OK, thanks.

And about mesh face vertices - I have a definition (from this post) that I believe has problems because the order of points for each mesh face is not consistent (i.e. always anti-clockwise or clockwise)

It fails for the same triangles that offset in the other direction above.

I presumed it was due to the face vertex order but I could be wrong!

All faces in your mesh are consistent, all counter-clockwise as expected.
With other thread’s script everything is fine (apparently) , to me.
You’ll need to be more specific about the problem and/or post a specific (and internalized) case …

The file is getting a bit wild! its the bit circled in red that I am trying to improve.

I know now that there is a problem with this approach because the mesh needs to be conical to work with straight (planar) struts.

I am trying to model what would happen if I made the mesh from rectangular section timber frame by making each mesh face and then adding the face angles on the timber edges.

And then look at whether its possible to modify the corners of the triangle so they fit together.

Maybe there is a better dome mesh than a geodesic dome that is suited to this manufacturing technique?