Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!samsung!umich!umeecs!zip!spencer
From: spencer@eecs.umich.edu (Spencer W. Thomas)
Newsgroups: comp.graphics
Subject: Re: rayshade and Utah Raster toolkit
Message-ID: <SPENCER.90Jun14233942@spline.eecs.umich.edu>
Date: 15 Jun 90 03:39:42 GMT
References: <10073@rouge.usl.edu>
Sender: news@zip.eecs.umich.edu
Organization: University of Michigan EECS Dept
Lines: 33
In-reply-to: pcb@gator.cacs.usl.edu's message of 14 Jun 90 23:32:17 GMT

> I recompiled with the RLE formats and ran several pictures.
> But when I put them on the SUN3 Color monitor, the 
> routine "to8" says   is not an RLE file.

Sigh.  I'm not sure what you're running rayshade on, but if it's a
little-endian machine (i.e., the low order byte in a word comes
first), and it's not a vax or pdp11, you will get this problem.  The
coming release (by the end of the month, really!) will totally fix
this situation on all machines, with no human intervention needed.

In svfb_global.h, find the line that says

#if !defined(vax) && !defined(pdp11)

and change it to

#ifndef LITTLE_ENDIAN

Then, if you are compiling the toolkit on a little endian machine
(vax, pdp11, x86, dec mips, etc.) add a line

#define LITTLE_ENDIAN 1

before this.

Sorry.  I hope this is the last time I need to post this message
(i.e., I hope it doesn't come up again before we get the new release
out.)



--
=Spencer (spencer@eecs.umich.edu)