Re: [Openrm-users] OpenRM/win9x: Illegal operations here, there, and everywhere!

[Wes B. <wb...@r3...>]

Hi Martin,

Once upon a time, I had an Elsa GeForce2 card that was possessed by an
evil demon. All OpenGL programs, regardless of OS, would wedge the 
machine after a period of time ranging from a few frames to several
minutes worth of operation. I've since thrown that card in the trash,
and have replaced it with a different nVidia card (I have several,
actually). It took much longer than I care to admit to discover that
this was a hardware problem - one never really expects the hardware
to be bad. (FYI, it was just that particular card, I hold no prejudice
against Elsa as I'm running with an Elsa 920 on the primary linux
dev box).

At that time, one of my "spare" cards was the exact same Matrox card
that you have. I really liked the Matrox card a lot. The OpenGL stuff
worked great for me, and my only complaint about the Matrox card is
that there is a bug in their mipmapping code. You can run the tmap2d
demo using the -mips flag to cause that to happen. I had no problems
with OpenGL acceleration on a WinNT box, and in fact, performance
was quite impressive. 

So, I suggest you remove the Matrox drivers, run in VGA mode and try
the demos again. If the problem goes away, your Matrox card may be
sick. If the problem persists, let us know and we'll go from there.

FYI, the only WinXX issues that have come up recently were the ones
where users with scrolling mice and clumsy fingers ;-) could make the
demos crash. Otherwise, make issues aside, the WinXX implementation has
been very solid. As of this date, we've tested OpenRM on Win95,
Win98, WinNT, Win2K and, finally, WinME. Gfx cards include a variety
of GeForce(1), GeForce2, GeForce3, Matrox G400, Matrox G200, and intel 
i810 & i815 gfx chipsets (not all WinXX combinations & gfx cards have 
been tested - one can expect only so much from dev folks on an Open 
Source project).

Thanks for the detailed system info in your report, BTW.

tx,
wes

Martin van den Nieuwelaar wrote:
> 
> Oh windows!#@!
> 
> Just tonight I installed rm140, rmdemo, jpeg-6b, and Pthreads all in
> the msvc++ 6.0 environment.  It seems to be working, sort of.  Some of
> the demo programs work as expected.  Others produce an illegal
> operation error as soon as a mouse control is activated (translate,
> rotate, zoom in/out etc.).  For example cones fails immediately after
> any mouse control.  tcube will run just fine for a few seconds of
> playing spin the cube :) before failing.  tfly seems to run just fine
> for at least a couple of minutes without any problems at all.  Just
> wondering if anyone has experienced these problems?
> 
> With all these inter-dependencies, listing the running configuration
> is tricky, but here goes...
> 
> Intel PIII-500.  192MB RAM.  Matrox G400 32MB dual-head.  Driver file
> version 4.12.01.1610.  win98 SE.  msvc++ 6.0 standard.  jpeg-6b
> compiled with the supplied jconfig.vc and makefile.vc.  PThreads
> pre-built ver 2001-07-12.  Demo programs makefile.w32 using
> pthreadVCE.dll.
> 
> I have heard the G400 OpenGL support is partly broken under Windows.
> At least Blender (the funky OpenGL modeller) apparently doesn't work
> because of this, or so I'm told.
> 
> Any thoughts appreciated,
> 
> -Martin (somewhat sleep deprived)
> 
> --
> Expert carrier network traffic analysis and visualisation
> http://www.gadgets.co.nz/products.shtml
> xenaphobia: The fear of being beaten to a
> pulp by a leather-clad, New Zealand woman
> 
> _______________________________________________
> Openrm-users mailing list
> Ope...@li...
> http://lists.sourceforge.net/lists/listinfo/openrm-users

-- 
//\/\\//\\//\\/\//\\/\\//\\//\\/\\/\//\\//\\//\\//\\//\/\\//\\//\\//\\
Wes Bethel                                          wb...@r3...
R3vis Corporation                                http://www.r3vis.com/
Phone: 415-898-0814                                 FAX: 415-898-2814
//\/\\//\\//\\/\//\\/\\//\\//\\/\\/\//\\//\\//\\//\\//\/\\//\\//\\//\\