Benedict

Catching up on this, suggestions would be: 1. leave as is (`nonconservative_terms`, `has_nonconservative_terms`, and `have_nonconservative_terms`) 2. use `nonconservative_terms` everywhere 3. use `has_nonservative_terms` everywhere 4. use `have_nonservative_terms` everywhere

> What do you think would be the best option, and why? 4., because - not 1 for consistency reasons - to me `have_nonconservative_vars` is a property of the equations...

CI now passes. Ready for comments @trixi-framework/developers

Maybe unrelated, but what was the idea behind https://github.com/trixi-framework/Trixi.jl/pull/576?

> Finally, am I correct that the auxiliary variables are recomputed, rather than interpolated, when the mesh is adapted? Does this give a different result than adding them to the...

> Maybe in the future we would like to be able to select which variables get recomputed and which interpolated/projected. 🤔 Thanks for these examples @amrueda! It seems clear to...

> I was also wondering, for problems without auxiliary variables, does this add any performance overhead? It doesn't look like it would be significant, but might be good to double...

> Thanks for these examples @amrueda! It seems clear to me that we do want to be able to choose either approach. I had a look. - The aux vars...

I addressed #2386, #2387. If we postpone the interpolation approach to a subsequent PR, do you think this is ready for the next round of reviews? @amrueda @tristanmontoya

With #2400 and the latest changes in `main` merged, this is now ready for the next round of reviews. Some comments: - The auxiliary variables could potentially be used everywhere...