William

While it sucks that there seems to be a bug, I'm glad that it's at least reproducible on other systems than mine. And thank you for the very quick response!...

Seeing the same thing. I'm at step 49 and still going strong!

I said that too early.. On step 50 it crashed with the same error :(

I've yet to try with the intel/mkl/impi approach, but installing the MPICH version ```bash # Name Version Build Channel dftbplus 21.2 mpi_mpich_h3469437_2 conda-forge ``` with conda still crashes at step...

Thanks for all the help! I might've have jumped the gun with my previous comment. With the conda installed MPICH version, it seems I'm able to converge structures consistently. I'll...

I can confirm that with `cargo build --all` there's no error. I'm on Ubuntu 22.04. Yet, even after building all examples, I'm still not able to run the specific example...

How's the work coming along with generating the spin constants for GFN2-xTB? I'm very interested in using GFN2-xTB through DFTB+ with a spin-polarized calculation for a molecule containing ruthenium

Wonderful! I'll try them out straightaway. Thank you!

Has there been any progress on this issue? If not, could I follow the approach of [`perses`](https://github.com/choderalab/perses/blob/main/perses/annihilation/relative.py) and scale the reciprocal part of the electrostatics like what is done there...

Bumping as I have the same problem.