Re: Slice Formats....

From: Peter H. Gien (71773.3150@compuserve.com)
Date: Sun Mar 03 1996 - 07:39:35 EET


Marshal Burns wrote:

> The thinking behind including the normal data was
probably to give the burden of computation to the program generating the
StL code, and reduce the computation needed in the slicing program.<

Computational burden of computing normal vectors from the right hand rule is
minimal. It is in fact simple and very fast. Cross product of any two oriented
sides of a triangle gives the desired result. As I recall, this involves two
multiplies and an addition for each component of the normal vector. For example,
50,000 facets would involve 50,000 * 6 multiplies and 50,000 * 3 additions.
Doing this on a 60MHz Pentium would take around 1/2 of a second. On my old
intel 80286/287 it would have taken around 2 seconds.

Given the above, I would have to disagree with your hypothesis. OK, would the
original perpertrator of the STL format please identify yourself!

Peter H. Gien
POGO International, Inc.
71773.3150@compuserve.com



This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:37:14 EEST