I've just been running a series of renderings in an animation and had
a couple of errors. I had interrupted the rendering and restarted it
using the Subset_Start_Frame and Subset_End_Frame commands in the INI
file. Pov-Ray Parses the scene ok and gets to 49% to 55% rendered when
it gives this error.
"The POV-Ray core rendering code caused an integer divide by zero
exception."
When I restarted POV-Ray and started to render from the same frame it
was fine. Actually I tried it with out shutting down POV-Ray and it
continued ok. This is intermittent. I don't know what else to include.
The scene is quite complex (for me) with two Poser Mesh2's and Scene
contains 2286 frame level objects; 0 infinite.

Regards
        Stephen

Post a reply to this message

Not a lot of response here :-{
Maybe if I rephrase my question. Why will PovRay give an "integer
divide by zero exception" one time and not another?

Regards
        Stephen

Post a reply to this message

In article <7augb0lum37lka9bau5f15pmf699ef2lah@4ax.com> , Stephen McAvoy 
<mca### [at] aolcom>  wrote:

> Not a lot of response here :-{

Well, you didn't provide a scene or any other kind of reproducible test
case.  You did not even state the version of POV-Ray you are using.  And you
said you already solved the problem.  Not much left to respond to then :-)

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

On Sat, 29 May 2004 16:19:59 +0200, "Thorsten Froehlich"
<tho### [at] trfde> wrote:

>Well, you didn't provide a scene or any other kind of reproducible test
>case.  You did not even state the version of POV-Ray you are using.  And you
>said you already solved the problem.  Not much left to respond to then :-)

Oh, I don't know about that, Thorsten.
Someone could have said something along the lines of "PovRay will give
an integer divide by zero exception when XXX" 
Seriously though I did not provide a scene for two reasons;
One I want to enter the animation in the current IRTC and I am loath
to preview it as any impact it may have, will be lost. Not that I'm
secretive or think someone would copy it but it is not ready to be
shown yet.
Two the size. The Pov file is greater than 12,500 lines and 660k in
size, it has 2 Poser meshes 2 Meg each. 
I don't want to waste bandwidth on something that might or might not
show up. That is the problem with intermittent faults. IIRC it only
started to give an error when I used AA settings of 
Antialias_Threshold = 0.1
Antialias_Depth = 3
Sampling_Method = 2
But that may be a red herring. 
What will cause an integer divide by zero exception when rendering?

BTW I am using PovRay 3.5, AMD 1735 Mhz, Win XP, 512 Meg of RAM.
Rereading my original post I think I can see why you and others may
have thought that the problem was solved. My head was as thick as
mince in the neck of a bottle due to a bad cold:-}

Regards
        Stephen

Post a reply to this message

In article <g5chb014t2rts9t67g3ifl8qhh8ke2goao@4ax.com> , Stephen McAvoy 
<mca### [at] aolcom>  wrote:

>>Well, you didn't provide a scene or any other kind of reproducible test
>>case.  You did not even state the version of POV-Ray you are using.  And you
>>said you already solved the problem.  Not much left to respond to then :-)
>
> Oh, I don't know about that, Thorsten.
> Someone could have said something along the lines of "PovRay will give
> an integer divide by zero exception when XXX"

Obviously when there is an exception something is wrong with POV-Ray, so
nobody can tell you where because then that would have been fixed if it was
indeed an error in POV-Ray!  However, if you say it went away, it is not all
that unlikely these days (because the structures are tiny, and personal
computers are not designed to be fault tolerant) that some bit in memory
just flipped.  To keep billions of tiny capacitors readable isn't as trivial
as one might think, and it can happen with a certain probability.  Also it
is far less than being struck by lightening, it happens so frequently server
memory modules come with error-correction bits built-in.  No kidding!

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

On Sat, 29 May 2004 19:38:19 +0200, "Thorsten Froehlich"
<tho### [at] trfde> wrote:

>To keep billions of tiny capacitors readable

Capacitors? What happened to those little iron tori? :-}
Thanks for the explanation. So what you are saying is these things
happen and there is not a lot that can be done about it?
Fortunately, the render went ok, for four days unattended. I just hope
that the next long render goes just as well.

Regards
        Stephen

Post a reply to this message