POV-Ray : Newsgroups : povray.beta-test : Unknown internal error. Server Time
31 Oct 2024 08:15:55 EDT (-0400)
  Unknown internal error. (Message 1 to 2 of 2)  
From: Sven Geier
Subject: Unknown internal error.
Date: 20 May 2009 14:55:00
Message: <web.4a1450e51b8bfa05b4449250@news.povray.org>
Upon encountering the error mentioned in the subject-line I'm vaguely unclear on
how to proceed - is there some way to force POV to "dump core" or output some
such diagnostic data that someone (certainly not me) could use to track down
what's going on?

I'm assuming "unknown" means that some error condition has been encountered that
the programmers have not provided for -- but I don't know what to look at to get
an idea of the "actual condition" (i.e. a function returning an undefined value
or a thread stopped responding etc).

Is there some way to get this kind of extra info?


Post a reply to this message

From: clipka
Subject: Re: Unknown internal error.
Date: 20 May 2009 15:55:01
Message: <web.4a145ee664a93a6cf708085d0@news.povray.org>
"Sven Geier" <.sven.at.sgeier.dot.net.nospamplease> wrote:
> Upon encountering the error mentioned in the subject-line I'm vaguely unclear on
> how to proceed - is there some way to force POV to "dump core" or output some
> such diagnostic data that someone (certainly not me) could use to track down
> what's going on?
>
> I'm assuming "unknown" means that some error condition has been encountered that
> the programmers have not provided for -- but I don't know what to look at to get
> an idea of the "actual condition" (i.e. a function returning an undefined value
> or a thread stopped responding etc).
>
> Is there some way to get this kind of extra info?

The most important thing in this context is - as always - your version of
POV-Ray, the operating system you used, whether you can reproduce the issue or
not, and such things.

If you can reproduce the issue, then the next helpful thing would be the scene
file you tried to render, parameters you used, etc. If you want to help the
developers as much as possible, you might try "boiling down" the scene as much
as possible to get some "minimal offending scene" (that's what they'd most
likely do first anyway).

If you can provide the developers with enough information that they can
reproduce the issue on their own systems, that's worth thousandfold more than
any "post mortem" information, because they'll have plenty of tools at their
disposal to examine the system in every little detail while the error is in the
process of happening, instead of just examining the mess it results in.

"Post mortem"-based debugging is typically used only as a last resort, when the
error has thwarted any and all attempts to observe it happening.


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.