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

CMake now takes care of extracting the major, minor and patch version of Python

You might have to adapt CMake flags when building OpenStructure to variables specified here:
https://cmake.org/cmake/help/latest/module/FindPython.html

SCHWED-4612: Set OpenGL GL preferences, new Python module name for sip in check if sip is installed.

This commit doesn't make OpenStructure work with Python 3. The goal of this
commit was to perform an automated port of the Python code and make it compile.
The performed steps:

- Edited CMakeLists.txt to search for Python with 3.6 as min version 3.6 is the
  Python version shipped by default with Ubuntu 18.04 LTS
- Add version 3.6 to cmake_support/FindPython.cmake
- Adapt setup_boost macro in cmake_support/OST.cmake to prefer versioned
  libraries and not first check for boost_python.so. In the example of
  Ubuntu 18.04, libboost_python.so is specific for Python 2 but
  libboost_python3.so is the one we want.
- apply the following command: 2to3-2.7 -n -w <OST_DIR>
- adapt base/pymod/wrap_base.cc, gui/pymod/wrap_gui.cc and
  gui/pymod/export_message_widget.cc as PyString functionalities do not exist
  anymore in the Python 3 interpreter (replaced by PyUnicode)
- adapt gui/src/python_shell/python_interpreter_worker.hh to resolve issue
  discussed in https://stackoverflow.com/questions/23068700/embedding-python3-in-qt-5
  Long story short: Qt does a typedef for "slots" which causes trouble with
  other headers that are pulled in from the Python interpreter

hasn't been used/tested for years

The reason is that we want to profit from latest security updates and
therefore not use a copy of the SQLite3 source code that becomes outdated.

ToDo: test on different Linux distributions / bc2 and adapt compile
instructions accordingly, also container recipes must be adapted.

What is NOT done yet:
- OpenGL context of QOpenGLWindow must be made
  active before calling any GL function in scene.cc. The wrong
  GL context is active otherwise.
- cmake polishing
- final grepping for QT4 leftovers

changes:

-GetPrincipalAxes in Vec3List
 Results with Eigen3 have been verified with 10 randomly picked pdb structures.
 Differences only occur with changing signs of the vectors.

-CalcPrincipalAxis in BoundingBox
 No difference in algorithm, the SelfAdjointEigenSolver no lives in
 Eigenvalues. Does in principal the same thing as the function above,
 just using different functionality.

-SVD superposer
 Unit tests still run through with the changes introduced for Eigen3
 and also manual testing seems to work well.

-CartoonRenderer
 Principal axes calculation for rendering helices as cylinders had to be
 adapted. Rendering looks exactly the same...

-SuperposeFrames
 Not yet tested...

ToDo:
numeric.h and levenberg_marquardt.h in the img/alg module have to be
ported. This will be done in the next few days.