Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!lll-winken!decwrl!sgi!tarolli@riva.esd.sgi.com
From: tarolli@riva.esd.sgi.com (Gary Tarolli)
Newsgroups: comp.sys.sgi
Subject: Re: update: RECTRE problems
Summary: rectread answer (I hope)
Keywords: possible clues
Message-ID: <45668@sgi.sgi.com>
Date: 4 Dec 89 17:02:19 GMT
References: <3022@uceng.UC.EDU>
Sender: tarolli@riva.esd.sgi.com
Organization: Silicon Graphics, Inc., Mountain View, CA
Lines: 33

In article <3022@uceng.UC.EDU>, trohling@uceng.UC.EDU (tom rohling) writes:
> 
> A few days back, I posted an article about RECTRE not appearing to be 
> functioning right.  Well, here's an update.
> 
	.
	.
	.

> 
> What !exactly! is the format of the result of RECTRE (details!)???
> 


If you say "man rectread" on 3.2, you get the following NOTES section in
the man page

NOTES
     These routines are available only in immediate mode.

     On IRIS-4D GT and GTX models, the upper 4 bits of the 16-bit values
     returned by rectread are undefined, and its performance will suffer if
     x2 - x1 + 1 is odd.

If you are running on a GT or GTX model (you didn't say, but from your
problem it sounds like you are), then
I think your problem is that the upper 4 bits of the 16 bit values are
indeed undefined, and are probably not 0!  This would make your comparison
to "3" fail even though the pixel is indeed color 3.  Hope this helps.


--
						Gary Tarolli