Vertex Array vs. Display Lists

I have a dynamic mesh, a heightfield, that’s broken up into several sections. Chances are that for many frames, some of these sections won’t change, and maybe one or two will. Would it be faster for me to make a display list for each section and then regenerate the DL when a section changes, or to just put it all into a vertex array thingy and that way modify vertex positions/orderings directly?

It really depends on the numbers. Some sort of display listing would probably benefit. You should probably split it into sections display list each section and regenerate the display list for modified sections. Exactly what the best granularity is should be a matter for experimentation. Other approaches may also be of interest, if you know in advance which parts of the terrain are likely to change you may want to keep these in system memory and dispatch with VBO or similar efficient mechanism.

Both share the disadvantage which is that OpenGL doesn’t want them to change their data.
I think vertex buffers are more flexible than display lists, so if you’d ask me I’d use them over display lists. :stuck_out_tongue:

You can also take a peek at vertex buffer objects…