Bálint Aradi

@Madebene I absolutely agree with you, the program should not silently truncate an exceeding line length (sorry for the current behavior). I've created a PR (#1066), which would for your...

Dear Xavier, thanks a lot for reporting the issue. I absolutely agree with you, that it is USPEX, which should be fixed. USPEX seems to use a fixed-format reader which...

Currently, the restarting procedure works as follows: * Set up a new calculation with the same parameters as the previous one. * Use the last geometry of the previous run...

I won't necessarily add calc_derivs to dptools at this stage. Currently, dptools is a package containing user convenience tools only and no developer tools. As for hsd-python: Yes, it definitely...

The `char` interface *extends* the intrinsic definition, so the intrinsic one is still available. As long as we only use it with our derived type, there should be no shadowing...

I converted this PR to draft in order to exclude from the list, if one filters the PRs with `-is:draft`... (If it is not a draft any more, feel free...

Did you try to change the solver (e.g. to QR)? Sometimes divideandconquer and relatively robust just stop.

Yes, unfortunately, QR is the slowest eigensolver within LAPACK/ScaLAPACK. But at the same time also the most robust one, and ususally does not stop unexpectedly. As an alternative, you may...

Thanks for reporting. You may have hit the issue reported in #940, which had been fixed already on the main branch. Please try your system with current main and let...

Yes, ideally DFTB+ should always return gracefully from errors. Unfortunately, this is not that straightforward in Fortran as it would be in other languages with exceptions. I would like to...