Member GEO_Detail::findInternalNormalAttribute () const

Remove internalN code path with the H11 viewport?

Member geomStringsMatch (const string &geom1, const string &geom2, bool contains=false)

Geometry name expressions are not yet supported.

Member GfVec2d::Normalize (double eps=GF_MIN_VECTOR_LENGTH)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec2f::Normalize (float eps=GF_MIN_VECTOR_LENGTH)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec2h::Normalize (GfHalf eps=0.001)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec3d::Normalize (double eps=GF_MIN_VECTOR_LENGTH)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec3f::Normalize (float eps=GF_MIN_VECTOR_LENGTH)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec3h::Normalize (GfHalf eps=0.001)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec4d::Normalize (double eps=GF_MIN_VECTOR_LENGTH)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec4f::Normalize (float eps=GF_MIN_VECTOR_LENGTH)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Member GfVec4h::Normalize (GfHalf eps=0.001)

This was fixed for bug 67777. This is a gcc64 optimizer bug. By tickling the code, it no longer tries to write into an illegal memory address (in the code section of memory).

Group Interface inputs of a node-graph.

describe scoped interfaces and fix bug/108940 to account for them.

File MultiResGrid.h

Add option to define the level of the input grid (currenlty 0) so as to allow for super-sampling.

Member onnxruntime::Node::MutableInputArgsCount ()

This should be removed in favor of a method that updates the input args and the count. Currently these operations are separate which is not a good setup.

Class onnxruntime::SchemaRegistryManager

We may add more consistency checks later.

Member openvdb::OPENVDB_VERSION_NAME::Grid< typename >::memUsage () const override

Add transform().memUsage()

Member openvdb::OPENVDB_VERSION_NAME::Grid< typename >::readBuffers (std::istream &, const CoordBBox &) override

Refactor this and the readBuffers() above once support for ABI 2 compatibility is dropped.

Member openvdb::OPENVDB_VERSION_NAME::io::Archive::writeHeader (std::ostream &, bool seekable) const

This method should not be const since it actually redefines the UUID!

Member openvdb::OPENVDB_VERSION_NAME::io::File::write (const GridCPtrVec &, const MetaMap &=MetaMap()) const override

GridPtrVec readAllGrids(const Name&)

Member openvdb::OPENVDB_VERSION_NAME::io::MaskCompress< ValueT, MaskT >::MaskCompress (const MaskT &valueMask, const MaskT &childMask, const ValueT *srcBuf, const ValueT &background)

Consider all values, not just inactive values?

Member openvdb::OPENVDB_VERSION_NAME::io::StreamMetadata::transferTo (std::ios_base &) const

Deprecate direct transfer; use StreamMetadata structs everywhere.

Member openvdb::OPENVDB_VERSION_NAME::io::writeCompressedValues (std::ostream &os, ValueT *srcBuf, Index srcCount, const MaskT &valueMask, const MaskT &childMask, bool toHalf)

Save the selection mask as long as most of the inactive values are one of two values?

Member openvdb::OPENVDB_VERSION_NAME::io::writeData< std::string > (std::ostream &os, const std::string *data, Index count, uint32_t)

add compression

Member openvdb::OPENVDB_VERSION_NAME::io::writeDataSize< std::string > (const std::string *data, Index count, uint32_t)

add compression

Member openvdb::OPENVDB_VERSION_NAME::math::CoordBBox::extents () const

deprecate - use dim instead

Class openvdb::OPENVDB_VERSION_NAME::math::MapAdapter< MapType, OpType, ResultType >

For now, the operator's result type must be specified explicitly, but eventually it should be possible, via traits, to derive the result type from the operator type.

Member openvdb::OPENVDB_VERSION_NAME::math::NonlinearFrustumMap::NonlinearFrustumMap (const Vec3d &position, const Vec3d &direction, const Vec3d &up, double aspect, double z_near, double depth, Coord::ValueType x_count, Coord::ValueType z_count)

check that depth > 0

check up.length > 0

check that direction dot up = 0

Member openvdb::OPENVDB_VERSION_NAME::math::operator+ (const std::string &s, bool)

These won't be needed if we eliminate StringGrids.

Class openvdb::OPENVDB_VERSION_NAME::math::Transform

Calculate an axis-aligned bounding box in index space from a bounding sphere in world space.

Member openvdb::OPENVDB_VERSION_NAME::MetaMap::MetaMap ()

this should really iterate over a map of Metadata::ConstPtrs

Class openvdb::OPENVDB_VERSION_NAME::points::VolumeTransfer< TreeT >

this specialization should avoid the probe

Member openvdb::OPENVDB_VERSION_NAME::tools::gridop::GridOperator< InGridT, MaskGridType, OutGridT, MapT, OperatorT, InterruptT >::process (bool threaded=true)

Can tiles be handled correctly without densification, or by densifying only to the width of the operator stencil?

Class openvdb::OPENVDB_VERSION_NAME::tools::LevelSetRayIntersector< GridT, SearchImplT, NodeLevel, RayT >

Add TrilinearSearchImpl, as an alternative to LinearSearchImpl, that performs analytical 3D trilinear intersection tests, i.e., solves cubic equations. This is slower but also more accurate than the 1D linear interpolation in LinearSearchImpl.

Member openvdb::OPENVDB_VERSION_NAME::tools::PointIndexLeafNode< T, Log2Dim >::readBuffers (std::istream &is, const CoordBBox &, bool fromHalf=false)

If any voxels were deactivated as a result of clipping in the call to BaseLeaf::readBuffers(), the point index list will need to be regenerated.

Member openvdb::OPENVDB_VERSION_NAME::tools::poisson::solveWithBoundaryConditionsAndPreconditioner (const TreeType &, const DomainTreeType &, const BoundaryOp &, math::pcg::State &, Interrupter &, bool staggered=false)

Is this really needed?

if (state.success) ... ?

Member openvdb::OPENVDB_VERSION_NAME::tools::QuadraticSampler::triquadraticInterpolation (ValueT(&data)[N][N][N], const Vec3R &uvw)

For vector types, interpolate over each component independently.

Member openvdb::OPENVDB_VERSION_NAME::tree::InternalNode< _ChildNodeType, Log2Dim >::IteratorBase< MaskOnIterator, InternalNode >

Make mask accessors public?

Member openvdb::OPENVDB_VERSION_NAME::tree::LeafNode< bool, Log2Dim >::IteratorBase< MaskOnIter, LeafNode >

Make mask accessors public?

Member openvdb::OPENVDB_VERSION_NAME::tree::LeafNode< bool, Log2Dim >::LeafNode (PartialCreate, const Coord &xyz, bool value=false, bool active=false)

For now, this is identical to the non-PartialCreate constructor. Consider modifying the Buffer class to allow it to be constructed without allocating a bitmask.

Member openvdb::OPENVDB_VERSION_NAME::tree::LeafNode< bool, Log2Dim >::LeafNode (const LeafNode< ValueT, Log2Dim > &other)

Consider using a value conversion functor passed as an argument instead.

Member openvdb::OPENVDB_VERSION_NAME::tree::LeafNode< T, Log2Dim >::IteratorBase< MaskOnIterator, LeafNode >

Make mask accessors public?

Member openvdb::OPENVDB_VERSION_NAME::tree::LeafNode< T, Log2Dim >::LeafNode (const LeafNode< OtherValueType, Log2Dim > &other)

Consider using a value conversion functor passed as an argument instead.

Member openvdb::OPENVDB_VERSION_NAME::tree::LeafNode< ValueMask, Log2Dim >::IteratorBase< MaskOnIter, LeafNode >

Make mask accessors public?

Member openvdb::OPENVDB_VERSION_NAME::tree::RootNodeCopyHelper< RootT, OtherRootT, true >::copyWithValueConversion (RootT &self, const OtherRootT &other)

Consider using a value conversion functor passed as an argument instead.

Member openvdb::OPENVDB_VERSION_NAME::tree::Tree< _RootNodeType >::print (std::ostream &os=std::cout, int verboseLevel=1) const override

Consider using hboost::io::ios_precision_saver instead.

not accurate for BoolTree (and probably should count tile values)

Member openvdb::OPENVDB_VERSION_NAME::tree::ValueAccessorImpl< TreeType, IsSafe, MutexT, IndexSequence >::insertNode (const Coord &xyz, NodeT &node)

deprecate?

Class SdfPrimSpec

Insert doc about references and inherits here.
Should have validate... methods for name, children, properties

Class UsdGeomConstraintTarget

Provide API for extracting prim or property path from a target's namespaced name.

Class UsdGeomModelAPI

CreatePayload()

Class UsdModelAPI

establish an _IsCompatible() override that returns IsModel()

GetModelInstanceName()

Member UsdShadeNodeGraph::CreateInput (const TfToken &name, const SdfValueTypeName &typeName) const

clarify error behavior if typeName does not match existing, defined attribute - should match UsdPrim::CreateAttribute - bug/108970

Class VariantAssign

Add support for variant assignments in graph traversal and string resolution.

Class Visibility

Add a Look::geomIsVisible method that computes the visibility between two geometries in the context of a specific Look.