View Single Post
Old 20th November 2007, 04:52 AM   #18
coldby
Senior Member
Professional user
 
Join Date: Jun 2006
Posts: 102
Default Wow...

Lisa, you certainly got me scared! However things begin to make some better sense now, thanks.
Still it seems to me that - in your example - in order to avoid co-planar edges between the two 'booleands' you should also rotate a trifle the copy: by merely enlarging it, the edges of #2 would stay parallel to #1's - and parallel lines cannot help being co-planar.
Generally speaking I tend to agree with you, though without being able to pinpoint the cause of the problem: apparently the current booleans detest whatever kind of "exact" alignment between the booleands: e.g. a sphere subtracted from a cube would produce a messy result if centered precisely on a vertex, but the mess would promptly disappear by misplacing either object by just an epsilon in any direction.
Why? I'll tell you as soon as I find a gypsy
coldby is offline   Reply With Quote