You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Then I load it into the panda3d app, add the nodes to the world, and the result is as on the screenshot above.
Although it's not possible to have identically named meshes in a single blend file, we can still have this case when importing models from other blend files via Blender's Link feature. Of course it's possible, but very hard to maintain globally unique names of the nodes across different blend files. Anyhow, if it can't be supported, I'd expect at least an error in case of interference rather than silently squashed nodes. So I assume it's a bug.
Notes
When I inspect the nodes in Python debugger they seem to be still separate in terms of references, shapes, etc.
Weird enough but even if I manually alter the repeating name property on nodes in Python, it doesn't help.
Not sure if it's relevant to meshes names only or something else as well (Materials, Objects, etc.)
I'm still not 100% sure if it's a blend2bam issue and not panda3d itself. But there are a few evidences that make me think it's blend2bam:
This doesn't happen when I convert and load models from 2 separate BAM files.
When I change the mesh name in Blender to something unique, the resulting level.bam slightly increases in size (assuming 2 bullet nodes getting generated instead of 1)
Example
Attaching a minimal example code with blend files.
Description
When I convert/load 2 models with different (but identically named) meshes, the second one gets a collision shape from the first one.
Use case
I have 3 separate blend files:
Cube
.Cube
.So I convert only the
level.blend
tolevel.bam
Then I load it into the panda3d app, add the nodes to the world, and the result is as on the screenshot above.
Although it's not possible to have identically named meshes in a single blend file, we can still have this case when importing models from other blend files via Blender's Link feature. Of course it's possible, but very hard to maintain globally unique names of the nodes across different blend files. Anyhow, if it can't be supported, I'd expect at least an error in case of interference rather than silently squashed nodes. So I assume it's a bug.
Notes
name
property on nodes in Python, it doesn't help.I'm still not 100% sure if it's a
blend2bam
issue and notpanda3d
itself. But there are a few evidences that make me think it'sblend2bam
:level.bam
slightly increases in size (assuming 2 bullet nodes getting generated instead of 1)Example
Attaching a minimal example code with blend files.
panda3d-name-bug.tar.gz
Environment
blend2bam 0.25.0
panda3d 1.10.14
Blender 4.1.1
Python 3.12.3
Thanks for taking a look :)
The text was updated successfully, but these errors were encountered: