Vladimir Tomov

Sorry @dylan-copeland, somehow I've missed this. I don't quite follow the use case. Calling `UpdateQuadratureData` always recomputes `dt_est`, so it won't be infinite after the call. Also, the quadrature data...

Yes, doing 1. `GetTimeStepEstimate` (this makes a call to `UpdateQuadratureData`) 2. `ResetTimeStepEstimate` 3. `Step` would break the method, but it also makes no sense to call a time integrator with...

Yes that would be good.

For gslib, see `miniapps/gslib/field-interp.cpp`.

I think @aaronskinner is already doing something very similar.

I'm not convinced that a new interface is needed. I achieve this in Laghos in very few lines: ``` Array ess_tdofs, ess_bdr(pmesh->bdr_attributes.Max()), dofs_list; for (int d = 0; d <...

I'm tracking down autotest changes in `remhos` after this merge. Happens only in remap mode (mesh moves) with `-pa`. Most likely in `DGTraceIntegrator::SetupPA`. In my case `rho` is `nullptr`, and...

The PA tests were already using/updating the geometric factors before this PR. I had the impression that this PR does not affect the use of the geometric factors, does it?...

What is the state of this? Do you still want to merge it @rcarson3?

@pazner can we remove the `in-review` label while waiting for the test case?