deal.II Newsletter #341
Rene Gassmoeller
Hello everyone!
This is deal.II newsletter #341.
It automatically reports recently merged features and discussions about the deal.II finite element library.
Below you find a list of recently proposed or merged features:
#19156: ROLVector: Check for ghost elements. (proposed by marcfehling)
#19154: Add missing definition of `new_isotropic_child_cell_vertices()` for hexes. (proposed by AndSte01)
#19153: Added back assertions to check whether the desired refinement method is viable. (proposed by AndSte01)
#19151: VTKWrappers utilities -- part VII -- data_to_dealii_vector (proposed by luca-heltai)
#19150: Untemplatify things that need not be templates. (proposed by bangerth; merged)
#19149: DataOut: Support non-contiguous distributed vectors via ReadWriteVector (proposed by mathsen)
#19148: Guard fegetexcept/fedisableexcpet correctly in ArborX tests (proposed by masterleinad; merged)
#19147: Replace some Asserts with compile-time checks (proposed by masterleinad; merged)
#19146: TpetraWrappers::SparsityPattern: Fall back to DynamicSparsityPattern if n_entries_per_row is zero (proposed by masterleinad)
#19145: VTK: Add missing symbols in interface module partition (proposed by masterleinad; merged)
#19144: Modules: Fix compiling without NOX (proposed by masterleinad; merged)
#19142: Fix two links to compress(). (proposed by drwells; merged)
#19141: Fix compilation when both TBB and taskflow are not available. (proposed by drwells; merged)
#19140: ReferenceCell: use more constexpr. (proposed by drwells; merged)
#19139: apply_to_subranges(): special case for small or 1 thread (proposed by tjhei; merged)
#19138: Make a function in ReferenceCell constexpr so that it can be used in assertions. (proposed by bangerth; merged)
#19137: Simplify connectivity.h. (proposed by bangerth; merged)
#19134: VTKWrappers utilities -- part VI -- read_all_data and load_vtk (proposed by luca-heltai)
#19133: VTKWrappers utilities -- part V -- GridTools::parallel_to_serial_vertex_indices (proposed by luca-heltai)
#19132: Fixed misspelling of "refinement" and minor refactorings. (proposed by AndSte01; merged)
#19130: No longer used pointers because we are no longer using polymorphism. (proposed by bangerth; merged)
#19128: Use `switch` statements instead of `else-if` for branching depending on refcell type in isotropic refinement. (proposed by AndSte01; merged)
#19127: Fix Trilinos module with Tpetra (proposed by masterleinad; merged)
#19126: Fix VTK module (proposed by masterleinad; merged)
#19125: gh-actions: Fix merge commit detection for CI. (proposed by marcfehling)
#19124: Add origin to detect merges utility (proposed by jpthiele)
#19123: Use RuntimeError in convert_header_file_to_interface_module_unit.py (proposed by masterleinad; merged)
#19122: Do not call `update_ghost_values` in ROLVector operations. (proposed by marcfehling; merged)
#19121: Remove CellType* hierarchy. (proposed by bangerth; merged)
#19119: CMake: autopilot: print build type and target name during build (proposed by tamiko; merged)
#19118: Move VTK version macros to source (proposed by masterleinad; merged)
#19117: Simplify ReferenceCell code, part 2. (proposed by bangerth; merged)
#19116: Simplify ReferenceCell code, part 1. (proposed by bangerth; merged)
#19115: Introduce ReferenceCell::face_and_line_to_cell_vertices(). (proposed by bangerth; merged)
#19114: Better document ReferenceCell::face_to_cell_vertices(). (proposed by bangerth; merged)
#19113: Better document ReferenceCell::face_to_cell_lines(). (proposed by bangerth; merged)
#19109: Enable more `Taskflow`-based parallelizations in parallel.h (proposed by QY-Shi)
#19108: Obtain information from ReferenceCell instead of CellType*. (proposed by bangerth; merged)
#19106: TrilinosWrapper: Deprecate overloads with defaulted numbers of entries per row (proposed by masterleinad)
#19105: RTreeFunctionalVisitor (proposed by luca-heltai; merged)
#19102: Avoid invalid operations in step-42 -- and do it better anyway! (proposed by bangerth; merged)
#19097: Introduce function that returns an iterator over the face indices of a certain type of a reference cell. (proposed by AndSte01; merged)
#19089: Do not allow modification of vectors with ghosts. (proposed by bangerth; merged)
#19080: SparseDirectMUMPS and LA::distributed::Vector (proposed by fdrmrc; merged)
#19062: VTKWrappers utilities -- part IV -- vtk_to_finite_element (proposed by luca-heltai; merged)
#19026: VTKWrappers utilities -- part III -- read_vertex_data (proposed by luca-heltai; merged)
#16950: Handle zero contour aligning with faces in MeshClassifier. (proposed by simonsticko; merged)
And this is a list of recently opened or closed discussions:
#19155 Regression tester regressed 95ef16 (opened and closed)
#19152 Regression tester regressed f05518 (opened)
#19143 `MappingQEulerian` in multi grid matrix free (opened)
#19136 Regression tester regressed cce24f (opened and closed)
#19135 Regression tester regressed 4f0116 (opened and closed)
#19131 Regression tester regressed ea81d7 (opened and closed)
#19129 Performance degradation from Taskflow (opened)
#19120 Regression tester regressed 07563e (opened and closed)
#19112 Issue with ReferenceCell::ReferenceCell::face_to_cell_lines() for quadrilaterals. (opened and closed)
#19111 Show info of debug or release mode when compiling (opened and closed)
#19110 Regression tester regressed 11d828 (opened)
#19107 Regression tester regressed 50a9b6 (opened and closed)
#18977 Performance degradation of `WorkStream` (`Taskflow` backend) since 9.7 (closed)
#16888 NonMatching::MeshClassifier: adequate treatment of face-conforming interface (closed)
A list of all major changes since the last release can be found at this website.
Thanks for being part of the community!
Let us know about questions, problems, bugs or just share your experience by writing to this mailing list, or by opening issues or pull requests on Github.
Additional information can be found at our official website.