home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!olivea!sgigate!sgi!fido!shogun.asd.sgi.com!erikl
- From: erikl@shogun.asd.sgi.com (Erik Lindholm)
- Newsgroups: comp.sys.sgi.graphics
- Subject: Re: RE concave polys
- Keywords: RE concave polys
- Message-ID: <1k2cs4INNg8u@fido.asd.sgi.com>
- Date: 26 Jan 93 03:57:56 GMT
- References: <alan.728007287@godzilla.cgl.citri.edu.au>
- Organization: Silicon Graphics, Inc., Mountain View, CA
- Lines: 33
- NNTP-Posting-Host: shogun.asd.sgi.com
-
- In article <alan.728007287@godzilla.cgl.citri.edu.au> alan@cgl.citri.edu.au (Alan Dorin) writes:
- >When drawing polys on the Reality Engine using concave(TRUE)
- >I get a few polys that appear to be missed out. (Ie. some
- >convex polys become invisible) The polys _are_ visible on
- >PI's and Indigo's though so the problem appears to be with
- >the RE.
- >
- >If I turn concave polys off, obviously the RE (and other machines)
- >screws up the truly concave polys. Anyone know of a bug that
- >would cause this and how to overcome it? (apart from removing the
- >concave polys)
- >
- >
- > Thanks,
- > Alan Dorin
- > (alan@godzilla.cgl.citri.edu.au)
-
- Does your concave polygons have any shared vertices? The concave decomposition
- code is very sensitive about illegal polygons and will abort early if any are
- detected. From the man page:
-
- Polygons whose edges intersect each other are never guaranteed to be
- drawn correctly.
-
- this includes repeating vertices. Each architecture tends to implement slightly
- different algorithms for concave polygons, and illegal polygons are good for
- detecting these differences.
-
- Erik
- --
- Of course I'm lying, .....but I think I love you. (Yello)
-
- .|\!/|. erikl@shogun.asd.sgi.com .|\!/|.
-