COLLADA vs Max Md2 Md3 or Milk

there is a special package deal with COLLADA - com.jmex.model.collada



others just a single class like:

AseToJme

MaxToJme

Md2ToJme

Md3ToJme

MilkToJme

ObjToJme



in jmetest package, the sample code loading COLLADA models (TestColladaLoadng) also different from those of Max, MD3, MD2 or Millk.



Does Collada format very different from 3Ds, MD3, MD2 or other 3D formats that It needs to take care  differently?

Collada is a little bit different in that it's a richer format with a larger spec. A single file that parsed a COLLADA file would be bigger than a single file that parses, say, a MD2 file.



However, I think the bigger difference is just one in approach between the programmers who wrote the different importers (or the same programmer, tackling these different problems).



Btw: Is the difference causing you a problem somehow, or are you just curious?

I think this is mostly a consistency issue. If you were writing a model editor, for example, you would need to have specific code for handing collada model loading.

I would like to have a simple rectangle with 2 animated bones in COLLADA (.dae) format exported by Maya (It seems to work), or 3ds, or anyother that worked in jmeImport, or TestColladaLoading (more especifictly). In blender3D, the COLLADA 4.1 exporter have some problems to work with JME, I did the "rectangle" simple test and nothing.



But I'm trying to discover the error by the export.py script of COLLADA 4.1. Then I have to verify the XML file of a skin and boned animated .dae (that worked in jme) to compare with my exported XML by hands.



Someone can help, please?