> On Wed, 3 Apr 2002 00:07:32 +0200, "JRG" <jrg### [at] hotmailcom> wrote:
> > That does not make it an infinite object.
> > I can read:
> > "scene contains 1 frame level objects; 0 infinite".
>
> IIRC this counting message consider bounding boxes so nothing strange (but of
> course can confuse).

You know that anything out of the bounding box is just clipped, don't you? So a
finite bounding_box _means_ a finite object.

>
> > Either an inverted object isn't infinite, or that message is wrong.
>
> Another test: make box, add inverse, hollow and media and tell me where media
> appear ?

Nowhere. I tried and it looked plain wrong.

> > BTW, I used difference many times, and my CSG objects never happened to become
> > infinte.
>
> You were lucky. Creation of bounding boxes for CSG isn't so simple to have one
> general rule. IIRC it depends on relation in space, volume, type of csg etc.
> Documentation says about bounding: "For difference and intersection operations
> this will hardly ever lead to an optimal bounding box. It's sometimes better
> (depending on the complexity of the CSG object) to have you place a bounding
> shape yourself using a bounded_by statement."

As a matter of fact I use bounded_by quite often. IIRC my coffee-percolator model
(mostly made of intersected planes) was about 300% faster with my manual bounding.

--
Jonathan.

Home: http://digilander.iol.it/jrgpov

Post a reply to this message