Hello Carsten, On 06.03.2013 16:50, Carsten Neumann wrote:
> > not sure, depends on where the commitChanges() calls (if any?) are. The > SimpleMaterial must have a chance to build the Chunks to populate the > parent ChunkMaterial. This is triggered by PrimeMaterial::changed() > (PrimeMaterial is ChunkMaterial's parent class) calling rebuildState(), > and the changed() calls are of course a side effect of commitChanges(). > commitChanges() is called and the rebuildState() function is also called. But I can't find the place where the parent ChunkMaterial chunk list is populated. In SimpleMaterial's rebuildState() the _materialChunk is updated and added to the _pState member. Additionally, the parent's addChunks(_pState) function is called which populates the _pState with the parent's chunks found in the _mfChunks container. However, the _mfChunks of the parent is never changed and that is actually what is evaluated in the MaterialChunkOverrideGroup renderEnter() function. Maybe the implementation of this function does make wrong assumptions about the chunks container content? As a side note I have discovered that I did have the very same problem back in 2010 and side stepped it then. http://www.mail-archive.com/[email protected]/msg12698.html Dirk did explain the reason but I must confess that it is not clear to me right now. Best, Johannes ------------------------------------------------------------------------------ Symantec Endpoint Protection 12 positioned as A LEADER in The Forrester Wave(TM): Endpoint Security, Q1 2013 and "remains a good choice" in the endpoint security space. For insight on selecting the right partner to tackle endpoint security challenges, access the full report. http://p.sf.net/sfu/symantec-dev2dev _______________________________________________ Opensg-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/opensg-users
