Louis Ponet

Thanks for investigating this! In my mwe, I meant that only the lines with `#error` behind them actually give an error. So the second line actually works as you also...

kshift didn't change much. The QE pseudos: Ni: US, pbesol, z valence = 18, 3s 3p 3d 4s 4p O: PAW, pbesol, z valence = 6, 2s 2p 3d Pr:...

well, inputs are in the first post, output for scf is: [scf.out](https://gist.github.com/louisponet/a84bad84f11525bfbc16a4be36bd1fe0#file-scf-out) The dftk output (DFTK convergence is much nicer than QE): ``` n Free energy Eₙ-Eₙ₋₁ ρout-ρin Magnet Diag...

I ran QE with the noinv = true, leading to the full 32 kpoints, but together with the lower cutoff of 45 it is pretty much the same as before....

I think 31.5s yes, The difference seems to be mainly in the first couple of iterations where DFTK does a lot more iterations to get the Hami diagonalized (18ish vs...

Okay so idk if it's a valid way of comparing, but if I take the mean seconds/diag iteration I get 2.433s in QE and 6.446s in DFTK. So it's not...

Seems like it: [gdkglcontext-quartz.c](https://github.com/GNOME/gtk/blob/4724a89022ef1bec93b1a42d4cf2fec7191ed712/gdk/quartz/gdkglcontext-quartz.c) That's a pity...

[It seems all relevant are supported?](https://developer.apple.com/opengl/OpenGL-Capabilities-Tables.pdf)

I think it would also be useful to have the schemas themselves in a json somewhere, so the other languages can use them too. Maybe there can then be a...

Hi, Thanks for looking into this. This is very weird indeed as the exec is certainly still there after this happens. I honestly think this has something to do with...