Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!tut.cis.ohio-state.edu!ucbvax!van-bc!ubc-cs!alberta!obed!steve
From: obed!steve@cs.UAlberta.CA (Stephen Samuel)
Newsgroups: comp.sys.sgi
Subject: Re: shademodel
Summary: But why doesn't it say something NICE?
Keywords: GL
Message-ID: <1990Feb19.064953.2125@cs.UAlberta.CA>
Date: 19 Feb 90 06:49:53 GMT
References: <2548@fs1.cam.nist.gov> <50750@sgi.sgi.com>
Organization: U. of Alberta: Biochemistry , Edmonton, AB
Lines: 26

In article <50750@sgi.sgi.com>, tarolli@riva.esd.sgi.com (Gary Tarolli) writes:
> In article <2548@fs1.cam.nist.gov>, blue@cam.nist.gov (Jim Blue) writes:
> > I just got my Personal Iris and was trying to run the examples in the
> > Graphics Library Programming Guide. The very first example bombed with
> > Segmentation Error. The offending line is
> > 	shademodel(FLAT);
> 
> The usual explanation for this is that you called shademodel() before winopen.
> Only a few of the window hints are legal to call before winopen.  The others
> --
> 						Gary Tarolli
Yep, this sounds like the right explanation, but what really bothers me is
WHY CAN'T WE GET MEANINGFUL ERROR MESSAGES????? 
  Something like 'Illegal call with no window open' would help users to
solve a problem without beating their head against a wall or calling the
hotline.  If it's possible to identify the graphics call that's blowing 
up, then this helps even more.

  Since your documentation doesn't always clearly state what can, or cannot
be called before a window open (hint, hint!), there is even more reason
to encourage reasonable error messages.

Stephen samuel 	  (userzxcv@ualtamts.bitnet   or  alberta!obed!steve)
-- 

Stephen samuel 	  (userzxcv@ualtamts.bitnet   or  alberta!obed!steve)