Revision: 44838
http://projects.blender.org/scm/viewvc.php?view=rev&root=bf-blender&revision=44838
Author: campbellbarton
Date: 2012-03-12 21:09:27 +0000 (Mon, 12 Mar 2012)
Log Message:
-----------
fix [#30529] BMesh: Wrong Indizes of Faces
problem was bow-tie quads would add opposite normals together and result in
zero vector which was used for projection.
Now is_quad_convex_v3() checks if quad contains 2 faces which point away from
eachother when split by either direction.
Theres another fix for this bug which can be done since creating the face can
use existing edges in the example given so it wont have to guess which order of
verts to use.
Modified Paths:
--------------
trunk/blender/source/blender/blenlib/intern/math_geom.c
Modified: trunk/blender/source/blender/blenlib/intern/math_geom.c
===================================================================
--- trunk/blender/source/blender/blenlib/intern/math_geom.c 2012-03-12
18:53:22 UTC (rev 44837)
+++ trunk/blender/source/blender/blenlib/intern/math_geom.c 2012-03-12
21:09:27 UTC (rev 44838)
@@ -3058,10 +3058,23 @@
{
float nor[3], nor1[3], nor2[3], vec[4][2];
int axis_a, axis_b;
-
+
/* define projection, do both trias apart, quad is undefined! */
+
+ /* strictly speaking this isn't necessarily convex,
+ * but when try normals point away from eachother
+ * we don't wan't to make that face */
+ normal_tri_v3(nor1, v2, v3, v4);
+ normal_tri_v3(nor2, v1, v2, v4);
+ if (UNLIKELY(dot_v3v3(nor1, nor2) < 0.0f)) {
+ return FALSE;
+ }
normal_tri_v3(nor1, v1, v2, v3);
normal_tri_v3(nor2, v1, v3, v4);
+ if (UNLIKELY(dot_v3v3(nor1, nor2) < 0.0f)) {
+ return FALSE;
+ }
+
add_v3_v3v3(nor, nor1, nor2);
axis_dominant_v3(&axis_a, &axis_b, nor);
@@ -3071,7 +3084,7 @@
vec[2][0]= v3[axis_a]; vec[2][1]= v3[axis_b];
vec[3][0]= v4[axis_a]; vec[3][1]= v4[axis_b];
-
+
/* linetests, the 2 diagonals have to instersect to be convex */
- return (isect_line_line_v2(vec[0], vec[2], vec[1], vec[3]) > 0) ? 1 : 0;
+ return (isect_line_line_v2(vec[0], vec[2], vec[1], vec[3]) > 0) ? TRUE
: FALSE;
}
_______________________________________________
Bf-blender-cvs mailing list
[email protected]
http://lists.blender.org/mailman/listinfo/bf-blender-cvs
