quite a strange problem: my software (using OpenGL) works just fine on many computers, but in some rare cases users get completely wrong graphics and/or the message “floating point division by zero” which seems to be caused by some OpenGL call. Unfortunately I can’t debug these errors because on all machines I have access to the program works properly!
I suggested updating the graphics card driver to those users, but the problem persists. Does anybody have a hint or an idea where to search the cause of the problem?
Thanks in advance!!!
I had a similar problem when I used Borland C Builder. Its C runtime by default doesnt disable the FPU exceptions. Under VisualC++ it worked fine.
You can disable them by using then _control87 function. Just look in your doc.
I also use Borland C++ Builder, so maybe your hint will suppress the floating point errors.
BUT as mentioned before, these errors sometimes come together with a strange graphics output…
One of my customers uses a Laptop with an Onboard Intel 830 M graphics chip. Maybe someone knows of incompatibilities with that kind of chip??