wanting to optimize performance of my app, I discovered that jME can make use of vertex buffer objects, but …
some older graphics cards only support mechanisms like display lists when it comes to caching geometry data.
Well, I know that display lists do only cache openGL commands in a way that the functions don't need to be called too often, but when I was programming with C++ and OpenGL in the past, they speeded up my apps quite a lot.
When vertex buffer objects are not used - do you use some kind of display lists with LWJGL ? (is it possible anyway?)
And a question in the same direction:
Do you use vertex arrays / normal arrays /… ?
Just asking …
upcoming functionality will allow you to lock static geometry data, doing so well compile it into a display list. Obviously, you wouldn't want to use display lists for non-static data.