Jesse Chan

Will do - traveling at the moment but should have time to look at this tonight

Thanks for adding this, @DanielDoehring! Sorry the review is a bit late. > I am not sure why we currently pass in `prandtl_number()` as a function, I believe this is...

Oops - I missed that. I don't think that has been added, but you could check the type of `mu` and create an anonymous function if it's a number? Only...

Yes, but previously we passed in mu as a value. We now expect that mu is callable so this is technically a breaking change. We can avoid this if we...

> > > For consistency with the other parabolic terms, see > > > > https://github.com/trixi-framework/Trixi.jl/blob/3c1762b20189cba6b27abaa88e1575dc495c5c73/src/equations/equations_parabolic.jl#L18-L19 > > > > > > > > and > > > > https://github.com/trixi-framework/Trixi.jl/blob/3c1762b20189cba6b27abaa88e1575dc495c5c73/src/equations/compressible_navier_stokes_1d.jl#L83...

I can confirm it crashes with the same error on 1 thread but runs on 5 threads. I'm running on an M2 ARM Mac with Julia 1.9.4.

I forgot that Julia only uses the C interface - thanks! I can't remember precisely which function changed, but I think it had to do with a function definition that...

It looks like not specifying `boundary_condition_periodic` for `:xneg, :xpos, :zneg, :zpos` fixes things. This is also consistent with https://github.com/jlchan/Trixi.jl/blob/b7be5856eba029a3d91257166be4ee62514ae0dd/examples/p4est_2d_dgsem/elixir_navierstokes_convergence.jl - I don't know why we didn't see that before...

Related: https://github.com/trixi-framework/Trixi.jl/issues/1510

See https://github.com/trixi-framework/paper-2022-robustness-entropy-projection/blob/6402bf40e13e50ce45fa3fd8852cb37e3092c9ae/julia_simulations/utilities.jl#L193-L217 for a 2D example