Veselin Dobrev

Discussion continued in #3205.

The version of SuperLU_Dist (v5 vs v6 and newer) is not auto-detected: when using v5, please set the flag `MFEM_USE_SUPERLU5` to `ON`.

Hi @yurivict, sorry for the late reply. I was not able to reproduce the issue. In the CMake output, what does the line containing `Found SuperLUDist:` look like in your...

In the CMake output, what does the line containing `Found SuperLUDist:` look like in your case?

This PR is probably not exactly ready for review (for merging in `master`), however, in order to get some feedback to @bensworth, we should add the label `ready-for-review` and start...

Hi @bensworth, can you please try to address the compilation errors? For example the errors from the `gitignore` test start here: https://github.com/mfem/mfem/pull/2474/checks?check_run_id=3362944018#step:9:89

An error estimator cannot depend on an unknown quantity (like E in the expression you have above). Typically, an error estimator will depend on the current discrete solution (like E_h...

I seems like there is some misunderstanding here: `dofs` is both input and output parameter here and its size should remain the same. We are simply converting "scalar" dofs to...

In @tzanio's instructions above there is an incorrect make flag: the MFEM GNU make build system does not use `CC` (which is typically the C compiler) but `CXX` which is...

By the way, if the problem is in building GLVis with Intel then you can probably just use `g++` to build a serial MFEM version just for use with GLVis...