un-modified, my own change didnt match up with the x3d spec.

by Michalis Kamburelis (kambi)

Normalize the axis before writing to avoid small values writing as a zero'd axis.

from dan grauer (kromar)

With change use_map_color_emission (which is for volumetrics only) to use_map_emit.

also ignore generated face-images (which have no file path).

   I also updated the Wiki page.
2. There was a small mistake introduced by the last commits. This has been 
   fixed.
3. Code cleaning

Blendphys

according to the "Style Guide for Python Code" 
(http://www.python.org/dev/peps/pep-0008/) and the recent review
(http://codereview.appspot.com/6815052). Some slight changes were done, too. 
Extensive tests have shown that the code is still well working.

Blendphys

Apparently internally SVG importer was using 90 points per blender unit,
not per inch. This made files importing with unexpectable dimensions
which could be really harmful if you're importing blueprints as references.

Now SVG importer is using 90 points per inch, which makes imported files
has correct dimensions in both Metric and Imperal unit systems.

However, there's one possible downside -- imported files would look
smaller in blender units system.

Blendphys

panel from the import/export addons. This means that the panel has had to be 
removed also from the PDB importer/exporter. This is what I have done.

Attention, this is a major change of the Atomic Blender PDB importer/exporter!

A lot of test today have shown, that the new version of the PDB 
importer/exporter has not lost its stability. In the next days, I will perform 
other stability tests.

Blendphys

Also in the PDB IO: relict from old times, which has been removed

Blendphys.

fix [#30740] N-panel addon toggle operators crash Blender after screen change (Fatal Python error: GC object already tracked)

rectangular leaves since bmesh.

use Tr as translucency and use reflection coords for 'map_refl' setting.

Some cosmetics ... .

When exporting structures from Blender into a PDB file it happened 
sometimes that the xyz positions of the atoms weren't correctly 
stored. Bug fixed.

import. I have no idea why that happened out of a sudden. Anyway, I removed
this bug.

In this situation, the script creates a dummy root bone... But code was buggy (and import of non-rotating bones was too).

Update to 2.64, and make the ticker slower to not eat CPUs.

Fix encoding error (please be carful with this - even though using non-ascii chars in code/comments is not a good idea, my bad). Thanks Irie Shinsuke for the report!