Repository navigation
Conversation
… IntVectorData primitive variables.
| IECOREALEMBIC_SPECIALISEGEOMPARAMTRAITS( Alembic::AbcGeom::IN3fGeomParam, IECore::V3fVectorData, IECore::GeometricData::Normal ) | ||
| IECOREALEMBIC_SPECIALISEGEOMPARAMTRAITS( Alembic::AbcGeom::IN3dGeomParam, IECore::V3dVectorData, IECore::GeometricData::Normal ) | ||
|
|
||
| IECOREALEMBIC_SPECIALISEGEOMPARAMTRAITS( Alembic::AbcGeom::IBoolGeomParam, IECore::IntVectorData, IECore::GeometricData::Numeric ) |
There was a problem hiding this comment.
Any reason this can't be BoolVectorData?
There was a problem hiding this comment.
IECoreRI/PrimitiveVariableList.cpp doesn't like BoolVectorData:
cortex/src/IECoreRI/PrimitiveVariableList.cpp
Line 189 in 6fa38ce
So for now, I've chose to just convert it to IntVectorData, to minimize changes.
There was a problem hiding this comment.
Hmm, I really think we should be using bools for this so that we can round-trip data correctly when we add an Alembic writer, and so that renderers that do support bools directly (Arnold for instance) can be properly supported. Do you think you could either use a DataPromoteOp somewhere in your pipeline on the way into RenderMan, or sit on this as a local change until we add a replacement RenderMan backend (it's due an overhaul since Gaffer added a new renderer API more suited to threading and IPR)?
|
I've added a note to #29 to handle bool data properly in IECoreAlembic. We'll be addressing that soon as part of Cortex 10, so I'm going to close this PR for now. Please feel free to make an updated PR targeting 9_maintenance with the change @johnhaddon suggested. |
3 participants
They get converted to IntVectorData primitive variables.