Wolfgang Bangerth

@anne-glerum Have you had time to think about @gassmoeller 's comment? The CI check that is failing is the one that uses the then-broken GMG functionality. If @anne-glerum is convinced...

For reference, the analysis in @YiminJin's paper only applies if you have that `eta(eps)` is a function of *only* `||eps||`. In that case, it is true that `d eta /...

Also for reference, if we wanted to address this, the function that does the work in question is in `base/utilities.cc`: ``` template double compute_spd_factor(const double eta, const SymmetricTensor &strain_rate, const...

Anisotropic viscosity still leads to a symmetric matrix. You have a term of the form `(eps_i, X eps_j)` where `X` is a symmetric rank-4 tensor that must satisfy `(eps_i, X...

As for the term added to the matrix, are we sure that it is actually correct? It might be worth going back to the paper in which it was proposed...

There has been substantial churn in the GMRES implementation in deal.II. Are you able to narrow down which patch to deal.II might have caused this? @kronbichler FYI -- perhaps related...

You'll have to merge the two commits (squash or fix-up) to ensure these large files are not in one commit and then removed in the next, leaving them as large...

It fails the indentation check with these diffs: ``` diff --git a/cookbooks/convection-box/tutorial-onset-of-convection/README.md b/cookbooks/convection-box/tutorial-onset-of-convection/README.md index a1d0535..1668690 100644 --- a/cookbooks/convection-box/tutorial-onset-of-convection/README.md +++ b/cookbooks/convection-box/tutorial-onset-of-convection/README.md @@ -29,11 +29,11 @@ If you run in your local...

The other failures seem to come from the fact that the root of the branch is at a point in time where we had the GMG failures.

What do we want to do here? I don't see a clear strategy for addressing the issue :-(