Fix #32329: inset produces incorrect result, epsilon was too small to handle

floating point error.
This commit is contained in:
Brecht Van Lommel
2012-08-31 16:10:13 +00:00
parent fe9b1c644f
commit 5ecff7a240

View File

@@ -283,7 +283,8 @@ void bmo_inset_exec(BMesh *bm, BMOperator *op)
* cross product between both face normals */
add_v3_v3v3(tvec, e_info_a->no, e_info_b->no);
if ((f_a == f_b) || compare_v3v3(f_a->no, f_b->no, 0.00001f)) {
/* epsilon increased to fix [#32329] */
if ((f_a == f_b) || compare_v3v3(f_a->no, f_b->no, 0.001f)) {
normalize_v3(tvec);
}
else {