deal.II Newsletter #253

[Rene Gassmoeller]

Hello everyone!

This is deal.II newsletter #253.
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:

#15193: Remove overhead from AlignedVector in FEFaceNormalEvaluationImpl (proposed by bergbauer)
#15192: Fix compilation of a test. (proposed by bangerth; merged)
#15191: Avoid variables named 'import'. (proposed by bangerth)
#15190: Adding wrappers for Ginkgo vectors (proposed by MarcelKoch)
#15189: Add a test with an irrecoverable failure of KINSOL. (proposed by bangerth)
#15188: Clean up after ourselves. (proposed by bangerth)
#15187: Catch and rethrow *all* exceptions. (proposed by bangerth)
#15186: Let NOX callbacks conform to our new standard for callback errors. (proposed by bangerth)
#15185: Weight kernel for multigrid: Simplify aliasing analysis for compiler (proposed by kronbichler; merged)
#15184: Add a ScopeExit class. (proposed by bangerth; merged)
#15183: Reduce operation count in polynomial evaluation (proposed by kronbichler; merged)
#15182: Reduce overhead in calls to tensor product value function (proposed by kronbichler; merged)
#15181: Avoid missing symbols and use constexpr functions instead (proposed by kronbichler; merged)
#15180: More NOX tests (proposed by bangerth; merged)
#15179: Fix up the single NOX test (proposed by bangerth; merged)
#15178: Let KINSOL callbacks conform to our new standard for callback errors. (proposed by bangerth; merged)
#15177: Make some variables in a few tests 'const'. (proposed by bangerth; merged)
#15176: Fix do_reinit of FEPointEvaluation for empty unit_points (proposed by bergbauer; merged)
#15173: Add the PETSc SNES version of step-77 as a test. (proposed by bangerth; merged)
#15171: Let doxygen resolve a citation in step-77. (proposed by bangerth; merged)
#15169: Move a function out of line. (proposed by bangerth; merged)
#15159: pass MPI_Comm by value (proposed by tjhei; merged)
#15154: make it possible to fill RPE with GT::DistributedComputePointLocation... (proposed by jh66637; merged)
#15144: Clarify the documentation of a reinit() function. (proposed by bangerth; merged)
#15141: Add two level transfer operator between non-nested levels (proposed by fdrmrc; merged)
#15138: Implement alternative rules what quadrature/mapping object to use in FEInterfaceValues::reinit(). (proposed by bangerth; merged)
#15132: Add 'requires' clauses to PETSc solvers. (proposed by bangerth; merged)
#15128: In step-77, also talk about the other nonlinear solver packages we support. (proposed by bangerth; merged)

And this is a list of recently opened or closed discussions:

#15096 Develop more tiebreakers for FEInterfaceValues::reinit() for different kinds of elements. (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.