roystgnr

> I think those nonsparse failures are likely real failures After half a day of trying to figure this out: they're worse than real failures. Real failures mean I wrote...

> I guess the whole point of this PR is the exodus output of these LMs Admittedly I do feel like I've gone down the rabbit hole here. I haven't...

Okay, I get it now. After making the orientation of lambda depend on geometry independent of numbering, I now get consistent lambda values on all side elements, and I was...

> do you think we should put in exception throwing for operations like TypeVector::norm if all coordinates are 0? Thinking about it, I can't avoid the conclusion that at some...

norm((x,y,z)) in 3D is a cone surface in 4D ... it's easy to picture in 2D, with a 3D cone. If you look at the directional derivative at 0, d(norm)/d(direction)...

> Also running Ubuntu 17.04 with the same issue, can confirm that @Octav14n's fix works. Works for me too. I now suspect this was the cause of the "Content servers...

Replaced by #3426

VTK_Group_MPI should be the suggested flag with older versions of VTK. I guess at this point in a configuration we have the vtk header, so we might detect version from...

Any chance the changes here are responsible for the new libMesh MOOSE-test CI failures? E.g. https://civet.inl.gov/job/2127543/ - 3 of our recipes are giving "unknown dependency" failures from the non-skipped `misc/signal_handler.test_signal_*recover*`...

We have a third party working on polygonal element support, though I'm not sure how much progress they're making. I don't think anyone's done anything with libMesh towards polyhedra. I...