Quote: "The limit is not 2^32, it is 2^16"
After i posted that and went to sleep i realised the mistake. The limit is 2^16.
Quote: "You can get 32 bit index buffers but some graphics cards don't support them. There is no reason why you should need that many vertices in a single mesh anyway, it is only going to be bad for performance. You should split them up into limbs or objects instead."
All graphic card manufactured in the last decade support 32 bit index buffers. And if you don't have that kind of a card you probably aren't using dbpro (because it's slow). I don't see how can it be bad for performance. When you have more limbs you need more draw calls to render the same amount of triangles. More limbs also means more vertex buffers->redundancy. More limbs also means more world matrices to calculate->this all slows things down.
Splitting one index buffer into many is slow and complex process.
Quote: "Anyway, before someone trots off to change the code to allow 32 bit indices, remember that doing so will not also break compatibility as noted above by Diggsey, but will also disable every plug-in that accesses the index buffer, including the official plug-ins. Y'know the ones - sparky's, loads of my own, darkphysics and ODE, and likely dark shader too, plus others."
Do you know that for a fact?
When you load object in x format with more than 22000 triangles dbpro doens't complain and creates 32bit index buffers. I haven't heard that anyone had problem with that king of objects. And i don't think that any of that plugins actually accesses index buffers - more likely they access index array so changing index buffer to 32 bit should be a problem.
Why do i need this? I'm creating a modeling program and i want to be able to import meshes from various formats without the need to split them. I can change that my self and compile my version of dbpro but i want that objects made in my program can be loaded with standard version of dbpro so it would be cool i you (IanM) make that changed for the official release of dbpro.
"After you finish the first 90% of a project, you have to finish the other 90%." - Michael Abrash
