Allocating ~8000 VBOs causes python to exit with the message "Fatal Python error: deallocating None" when PyOpenGL-accelerate is installed. It's easy to hit this limit with rapidly changing content, e.g. incoming experimental data. Uninstalling PyOpenGL-accelerate avoids this crash, and the code seems to work fine in that case. This is a reasonable workaround for the moment, but it would be nice to be able to use PyOpenGL-accelerate in the long term.
System: Ubuntu 9.10, AMD_64, PyOpenGL 3.0.1b1, Intel X3100 graphics
Here's a minimal reproduction case:
from OpenGL.arrays import vbo
import numpy as np
data = np.arange(1000).astype(np.float32)
for i in range(1000000):
new_vbo = vbo.VBO(data)
# optional: new_vbo.delete
Log in to post a comment.