comments on the solutions to the new problems:
----------------------------------------------

While most of the solutions for the first 5 supplied problems turned out
to be quite nice, the 5 new problems seem to be good examples for the
obvious limits of our domain (see README).

We obviously (well, obviously if you look at the solution with some
piano playing knowledge) have a lack of operators (moves) which leads
to a much to high cost-evaluation of the human solution and to 'funny'
computer solutions.

The 'solution' to 'French' (failure !) demonstrates, that you can run into
trouble using REJECT in SCRs:
  After playing the D2  5 times we reject to JUMP to the next note (ef2).
  As another SCR prefers to start playing with the thumb, we run into a dead
  end...
  (However, we should backtrack and start with a different finger. Seems like
   there is some trouble in the domain still... By now it is too late to work
   on this, I suppose it is an error in some SCRs)