1 1. Replace SUMSL with LBFGS everywhere (currently only minim_ecart)
2 2. Parallelize the transformation of Cartesian derivatives to backbone-angle
3 derivatives (cart2int.F)
6 5. Fourbody interactions - eliminate pp-contact etc. storage (currently 4body
7 switched off by default, -DFOURBODY flag switches on).
8 6. Fix LBFGS - the linesearch (search) subroutine does not handle difficult
9 cases when changing variables can result in nans.