[Bf-python] Collada Importer - Issue with Mesh Vertex Normals

[Gaia Clary]

Hi, Khalibloo

The collada format allows an optimization that only stores unique normal 
vectors. Then later the normal vectors can be referenced by using 
indexes. Also there can be more than one normal per vertex since Vertex 
normals are defined per face AND per vertex. Because of this the number 
of normals per vertex can be anything from 0 up to the number of faces 
which share the vertex.

If you are sure that you found a bug, then please file a bug report and 
include a demo file that shows exactly where the data goes wrong. Then 
we will take care of this of course.

BTW: If we talk about Blender's built in collada module, then please 
note that this is a functionality that was created in C++ and it uses 
the OpenCollada library.

cheers,
Gaia

On 22.06.2016 21:06, Khalifa Lame wrote:
> Hello. I recently did some work on collada files exported from 
> blender. From what I understood from the collada wiki from Khronos, 
> the mesh normals xml node in the collada file should contain a trio of 
> number values for each vertex. In other words, the mesh normals should 
> have exactly the same number of entries as the vertex positions.
>
> This isn't always the case, however, when you inspect collada files 
> exported from blender. Sometimes, there are more normals than 
> vertices; sometimes, there are less. They hardly ever match.
>
> Is this a bug on blender's part? or is this intentional? if so, could 
> someone please help me understand the logic behind it and how to read 
> the normals reliably?
>
> This doesn't seem like a python question specifically, but if it turns 
> out to be a bug, I'll have to write a python script to modify that 
> behavior.
>
> -- 
> khalibloo®
>
>
>
> _______________________________________________
> Bf-python mailing list
> Bf-python at blender.org
> https://lists.blender.org/mailman/listinfo/bf-python

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.blender.org/pipermail/bf-python/attachments/20160622/bcd5ab90/attachment.html>