Howdy. Since there's been a lot of changes with how CL handles image data, I have a query.

In the old versions of CL, to use and re-use surfaces, I had a simple std::vector class that loaded all my game data (I called it a surface buffer), then I copied from that list to make objects, then destroyed those copies but the original allocation remained in the surface buffer. This way, I always had a loaded copy of the image in memory, so no slow-down or reduncancy for reloading on the fly. It also made clearing memory a little neater.

With the texture packing and grid clipping routines now, and based on my pretty poor overall underatanding of the way OpenGL handles uploading textures to the hardware, is this still necessary, or is my surface buffer idea obsolete?

I figure it is necessary to some degree, but I'm just curious. If anyone has a better idea on how to handle images and memory allocation, I'm all ears =D