Revision: 73068
http://sourceforge.net/p/brlcad/code/73068
Author: starseeker
Date: 2019-05-15 14:39:49 +0000 (Wed, 15 May 2019)
Log Message:
-----------
Saw a weird case where three edge points ended up triangulated and
non-degenerate in area, but the triangle ended up perpendicular to the overall
mesh in that area. Was topologically closed, but that's not what we want if it
can be avoided...
Modified Paths:
--------------
brlcad/trunk/src/libbrep/cdt.cpp
Modified: brlcad/trunk/src/libbrep/cdt.cpp
===================================================================
--- brlcad/trunk/src/libbrep/cdt.cpp 2019-05-15 14:31:37 UTC (rev 73067)
+++ brlcad/trunk/src/libbrep/cdt.cpp 2019-05-15 14:39:49 UTC (rev 73068)
@@ -3463,6 +3463,15 @@
}
}
+ // TODO - it's even possible in principle for a triangulation to form a
non-zero,
+ // non degenerate face from 3 edge points that will "close" the solid but
result
+ // in a face with all of its vertices normals pointing in a very strange
direction
+ // compared to the surface normals at those points (not to mention looking
"wrong"
+ // in that the "proper" surface of the mesh will look like it is flattened
in the
+ // area of that triangle. If we can recognize these cases and splitting
them will
+ // improve the situation, we want to handle them just as we would a zero
area
+ // triangle...
+
/* For the non-zero-area involved triangles, we need to build new polygons
* from each triangle and the "orphaned" points along its edge(s). We then
* re-tessellate in the triangle's parametric space.
This was sent by the SourceForge.net collaborative development platform, the
world's largest Open Source development site.
_______________________________________________
BRL-CAD Source Commits mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/brlcad-commits
