Jean-Paul Pelteret

> I think having a short meeting with all people interested would be great to see other opinions on this topic and see how we can set things up and...

Good idea to construct this list, @bangerth. Since I already moved the milestone for step-73, I took the liberty of shifting it to the "later" category in the list.

> Unrelated to the concrete patch here -- doesn't PETSc also wrap SuperLU and could use that solver? Trilinos has a wrapper for this ([Amesos2 docs](https://trilinos.github.io/amesos2.html)), so I'm sure that...

Partially. I only fixed the tests that you directed me towards, but there's still scope for improvement (I think, since I didn't look at the other tests).

Thanks for working towards optimising all things related to the AD types, @drwells. I'll do my best to give you feedback for both this and your other PRs on the...

This is still a WIP, so please do not review or merge this yet.

@dealii/dealii I've just finished this code-gallery example. It would be nice if someone could quickly skim over the README file to proof read it. Otherwise, I guess I'll merge it...

Thanks for reviewing this a few days ago @peterrum. I just wanted to drop by to say thanks, and that your comments have lead me to doing a bit more...

Currently WIP because I would like to generalise this to all `ad_type_code`s that we support (i.e. enable ADOL-C support and choose the AD library in the parameter file).

> However this constraints the choice between residual and energy AD purely based upon n_supported_derivative_levels meaning one can't use adolc_taped for residual linearization. You could get around this with a...