New Interface for GDOPT (General Dynamic Optimizer - Python Modeling Environment)

    - generate Code (e.g. just raw C-Code or use a Python AD-lib for C-Code gen)
        - with parameters
        - runtime parameters
        - exact hessians
        - exact sparsity
        - extended modeling (more generic M and r -> define x0 = m.addInitialState(x) / m.addFinalState(x))

    - defines sparsity and model structure

    - create yaml configuration (needed for OM and GDOPT interface), use yamlcpp for workflows and custom strategies

    - how to call a minimal simulation of it inside GDOPT (just from callbacks)?
        -> maybe create a simple ESDIRK(4) or Radau5 implementation :)
