OK - I have the fixes. I checked them into contrib/RTR; the files are
Here are the specific fixes I did:
Robert The Rebuilder wrote:- plLayerAnimation.read() encounters a NoneType for self.fTimeConvert
I added a check for None then assign it an instance of plAnimTimeConvert.
Robert The Rebuilder wrote:- plAnimTimeConvert has no getVersion()
I added a getVersion method and a version variable, as well as a default version constructor argument.
Robert The Rebuilder wrote:- plAnimTimeConvert.read() needs to attempt to read the messages that are implemented to date (see plCameraModifier1.read() for an example)
I borrowed code from plCameraModifier1.read() for reading in the messages, surrounded by a try/except that traps any errors in reading and throws an exception to abort the input (with an appropriate error message). This is necessary until we have all the possible message classes implemented.
Robert The Rebuilder wrote:- plLayerAnimation has no ToBlenderMTex() method
Added a stub method. Since plLayer.ToBlenderMTex() sets mtex.colfac and mtex.neg according to the layer's opacity and blend flags, it didn't seem to apply to a layer animation. So, this stub method does nothing except keeping the import going to the finish.
Robert The Rebuilder wrote:- plCameraBrain1_FirstPerson.read() claims that it has no attribute fOffset because plCameraBrain1_FirstPerson.__init__ needs to call plCameraBrain1_Avatar.__init___(), not plCameraBrain1._init__()
Made that fix.
Robert The Rebuilder wrote:- Imported visual objects do not appear in the blender scene - only physical objects
This is not true; the visual objects do appear in a blender scene named after the page; however, physical objects are placed in the default scene. I didn't address that bug in these fixes.
Could someone review these and check them into trunk for me, please?