Daniel Karrasch

@nanosoldier `runtests()`

@nanosoldier `runtests(["EulerAngles", "Groups", "Pyehtim", "DelayEmbeddings", "OptimalPortfolios", "Miter", "ThermodynamicIntegration", "Polynomials4ML", "IRKGaussLegendre", "OrdinaryDiffEqHighOrderRK", "OrdinaryDiffEqQPRK", "OrdinaryDiffEqFeagin", "OrdinaryDiffEqLowOrderRK", "OrdinaryDiffEqStabilizedRK", "OrdinaryDiffEqNordsieck", "OrdinaryDiffEqAdamsBashforthMoulton", "OrdinaryDiffEqSSPRK", "MCMCTempering", "OceanRobots", "MicroCanonicalHMC", "FMIImport", "BiochemicalAlgorithms", "OptimizationOptimJL", "Simpsons", "JumpProblemLibrary", "RealTimeAudioDiffEq", "GeneralizedSasakiNakamura", "VLBIPlots", "RobustBlindVerification",...

Seems like we can go ahead and remove it. I just wonder if this should really be backported.

Uff, this was supposed to supersede https://github.com/JuliaLang/julia/pull/55414, I guess?

> I'll rebase this to resolve the conflicts. I have done that already. In the resolution, I kept the zero-size early drop-out.

`ldiv!` for `BunchKaufman` is implemented in LAPACK. One would need to check if LAPACK also provides right-solves. Alternatively, this may fix it: ```julia /(A::StridedMatrix, B::BunchKaufman) = copy((B' \ A')') ```...

LinearMaps.jl has support for all kinds of lazy "matrix" manipulations: addition/subtraction, scalar/operator multiplication, Kronecker products, horizontal/vertical/diagonal concatenation. You can use any `AbstractMatrix` types as building blocks.

The reason is that that `Symmetric` wrappers only look at one, say the upper, triangle, and as for elements of the lower triangle, we pick elements from the upper and...

> On a second thought, the easiest solution would be for `promote_op(*,AbstractMatrix,AbstractMatrix)` to return `AbstractMatrix`. Good question why that is not the case.

Probably since https://github.com/JuliaLang/julia/pull/53573.