Versions are updated to what the mighty Schwede lab uses in production

Compilation was broken for years when IMG has been disabled.
IMG is now always built.

If fault_tolerant is enabled, it might happily read a cif file as PDB file.
As a consequence, there are no residues in the returned entity.

QT5 may cause hard to interpret import errors on older Linux kernels:
https://bbs.archlinux.org/viewtopic.php?id=232682

Otherwise it will fail if you provide a non-absolute working directory

Can either be pushed to standalone internal file format
or serve as starting point for MMTF / binaryCIF implementation
in OST. The stored data is kind of similar...

same as commit 69f90c46

makes it easier to re-use elsewhere

The test_Interpolate unit test has been reported to fail when executed
on different platforms (arm64 in this case).

The test_simulation_energy_calculations unit test compares the potential
energy of crambin after 200 steps steepest descent with a reference value.
Depending on OpenMM version / hardware, we might get slight differences
that propagate through the minimization.