roystgnr

> But I gather that's not possible as of now. It's totally possible! The trouble is that, either you sync the changes to ghosted elements (i.e. all elements, on a...

I haven't properly profiled, just looked at libMesh PerfLog and glanced at MOOSE output, attached here: [replicated16.txt](https://github.com/user-attachments/files/15998022/replicated16.txt)

Running CI for Draft PRs is definitely the behavior I'd intend. I write ~~too much~~ occasional code that I want to run through the whole CI gauntlet even when I...

I'd rather not use `PetscCall` ever, personally. `LIBMESH_CHKERR` doesn't avoid using `CHKERR` not because the latter is a recent invention, but because throwing an exception that can either be caught...

Is this ready to review? I've been busy, and currently the recent force pushes make me cautious, but if I don't get to it in between vacation prep work tomorrow...

> This is obviously not true if we compute orientation based on node ids, which only depend on how we happen to count nodes. And on how we happen to...

Oh that's beautiful. I'm of the general opinion that we should add as many orders as floating point arithmetic lets us get away with; in FEInterface I set `unlimited =...

This is really astonishing. "Refine a second-order mesh" is a *really* common thing to do. Let me see if I can replicate.

Even in MOOSE it seemed like "go second order, refine" would be both straightforward (you call one libMesh function, then another, and order doesn't matter) and frequently used. I figured...

We do build a `QUADSHELL4` if the ExodusII file specifies a `SHELL4` rather than a `QUAD4`, and that would p-refine to a `QUADSHELL8`.