Susi Lehtola

Well, you would just need to call the Fock matrix builder with the given density matrices; then you will have orbitals that you can feed into Psi4. This is exactly...

Yup

> I've got a branch that's only two days removed from master, and its smoke tests are clean. There is a problem with psi4 and qcng 0.28, but I'm sure...

The real problem here is that both the old and the new optking crash because they are unable to find out a coordinate system for the molecule.

We encountered this issue in metadynamics on the vanillin molecule with the coordinates ``` $coord -3.14393748772232 3.33813352345031 -0.00014119336956 o 7.86002766715000 -1.65701448465138 0.00040621434150 o 5.74049590710396 -2.49527551648547 -0.00050479923313 c 3.42304695453588 -0.98774709392443 -0.00010485496334...

For instance, running ``` xtb --omd --input metadyn.inp coord ``` on my laptop with `xtb-6.5.1-1.fc37.x86_64` I got the following result ``` time (ps) Ekin T Etot 0 0.00 0.00000 0.0613...

@evaleev yes, progman.pdf is generated. Note that the issue is not that latex2html does not exist, but that the make rule appears to be broken.. I think latex2html works on...

Ugh, now in 2.8.2 I do get progman.pdf but I don't get the classdoc which was generated in 2.8.0...

Actually, looks like the classdoc wasn't generated in 2.8.0 either!

> Perhaps the SONAME of qcint can be changed so that pyscf built against libcint cannot dynamically load qcint? Yes, definitely! This would be the correct solution.