well, a temporary buffer doesn't help. I use one buffer only and depending on whether there's half float uvs in there or shorts you choose (function) HF_UV or (function) WordUV from the dropdown combobox. The "knowledge" how to treat the values in the buffer is contained in those function. (It was a bad architectural design decision, I know, but it can't be corrected by simple means, imho.)Bigchillghost wrote:So you do all the works in every single function combination? Why not just read the vertices/UVs data to a temporary buffer and convert/copy them to the destination buffer? Then you need to process only one buffer if the flipping UVs option is chosen.
CF is combined face indices, not implemented.BTW I'm still confused about the UI. What are the w/64k / w/8k and the grey CF buttons for?
The w/x is active for WordUV and ShortAll only as you may have noticed, it's just the divider for word (short) uvs because there were formats where the uv map appeared "zoomed out".
It's for bigger meshes where the submesh borders are hard to find.Also, I noticed that the exported mesh is grouped per 500 triangles, is it because the face buffer is that big, or that it's a designed feature? I wish I don't have to blocked them out every time.
In blender wavefront importer there's the option to load Split by Group.
If you uncheck that feature it's remains for all mehses loaded afterwards still you check it again or restart blender.
(I didn't think about other 3D apps, so sorry for that.)