On 28.03.10 19:36, clipka wrote:
>> The geometry problems are the result of the bounding going wrong.
> No, definitely not - geometry code *MUST NOT* rely on any undesired
> surfaces being eliminated by bounding - if only because rotating the
> object may grow the bounding box.
I think there is some miscommunication: I am not saying that the bounding
and the object depend on each other in any unusual way. I am just saying
that for certain cases with cubic splines the bounding will be too small.
Post a reply to this message