Previously, it appeared visually that only one material was applied to the
mesh. Appearances can be deceiving. We were actually exporting all
materials and such correctly. However, because we were using a dyanmic
type with class attributes to store per-material data... All materials
shared all data. So, all materials rendered all vertices. You saw visually
the one last material rendered. Fun, eh?
This is done using the new checkboxes on the page list under the World
panel. Seemed like the best way to do it within our format. Perhaps the
rest of the exporter options should move to that panel at some point.
- Applying scale to *all* objects in the scene is now a pre-export step.
- Make sure we always export the physical's coordinates--this matches
Plasma's behavior.
Cyan calculates the level sizes wrongly such that a 1x1 DXT5 mip level
only has one block for alpha and therefore lacks the color block. This can
cause Blender to crash with an access violation.
This commit adds a footstep region modifier that exports nothing at the
moment. This is just a test case for logic node generation. If you want to
bake footstep regions into logic nodes, select an object and execute
`bpy.ops.object.plasma_logicwiz()`. :)
This includes changes to the light baking code to ensure that we don't
bake runtime lights. This code has several places it could be optimized in
the future when we have larger ages to test against.
This should make the average age look a little bit better. I hope... If
you want to get rid of this crappy lighting, create a vertex color layer.
Or, better yet, add a stupid lightmap modifier!
Still lots of considerations:
- Need to make faux lightgroups to avoid baking Plasma RT lights
- Still need to bake vertex colors
- Still need to export Plasma RT lights (D'oh!)
But, this is definitely a step in the right direction!
... This can happen if you save a blend file with a downstream codebase
with some newfangled modifier, then try to export with an upstream plugin
that does not have said modifier.
We now harvest information about texture images throughout the export
process and don't pull the trigger until we're totally done. There are
still some issues with regard to confusing UI, but we'll get to it.
Yeah, korlib was a bit faster than this implementation, but, honestly,
having a second python module made life more difficult than it needed to
be. This fixes that.
