Something I don’t understand.
After you run _UnweldEdge, shouldn’t the edges you unwelded be detected as “naked edges”? You can see them if you have “Mesh Edge Settings” (under “Display Modes”->“Objects”) set to be thicker and/or a different color.
The file I posted above from Blender looks like all the edges are unwelded (didn’t notice it before I changed the “Mesh Edge Settings”). The edges all appear to be connected when I look at it in Blender - strange.
I’ll look at this further tomorrow - out of time for now.
–Edit-- next day - further investigation:
There are definitely differences in how meshes are handled in Rhino and Blender.
In Rhino V6 I created a mesh sphere, copied it and ran _Unweld with Angle Tolerance = 0 and Modify Normals unchecked.
The unwelded mesh looks different because of display mode settings:
I sub-object selected a face on each mesh and moved it - you can see that the faces are all still connected:
I exported this file to obj - I know there is a setting for Vertex Welding on export, I have it set to “Unmodified”:
When I open the file in Blender and examine the unwelded mesh all the edges show as Non Manifold (Boundaries).
If I grab and move a face, you can see the faces on the unwelded mesh are not connected:
You can also see the effect if you apply a Subdivision Surface modifier - all the disconnected edges shrink away from each other:
The same effect can be seen in ZBrush if you move a face:
Or if you divide the mesh:
Back to the original topic of this post: If you have to unweld edges in Rhino to be able to have disconnected UV islands, or as you called it “introduce texture discontinuity on welded areas of a mesh in OBJ export” - the disconnected edges cause undesired side effects in Blender and ZBrush. In both of these programs you can UV map a mesh with disconnected UV islands without disconnecting edges.
Apologies for being so long winded, trying to explain clearly.