Concrete TODO:

- Run group computations with Malcev collector. 

- integrate code from symbolic.g
  How can we speed up the computation of the star polynomials ? 

- investigate why the computation of the log,exp polynomials
  is so  fast, once we have the star polys. 
  For example in F_2,8 using the symbolic Lie bracket is very slow. 
  But it seems that using the star pols for computing log and exp
  does not take much more time.

- Long term aim:
  The user comes along with a pc-group G.
  In a first step a good polycyclic sequence of G is computed. 
  Second the Mal'cev correspondence is set up and everything else
  that is needed for the fast multiplication. 
  Then collection can be performed using this setup. 

 
