Make them a parameter instead.  This allows to switch more easily between
geodesic finite elements and projected finite elements.

It is still no function in the dune-functions sense, but at least the
basis parameter needs to be a dune-functions basis now, rather than
a deprecated dune-fufem basis.

I intend to move as much towards the dune-functions interface for functions,
and this is a first step.

This implements a global GFE function (inheriting from VirtualGridViewFunction).
However, the return value of this function is _not_ a manifold class.  Rather,
the manifold is embedded into a Euclidean space.  Hence this class is formally
simply vector valued, and can be used in error measurement routines and such.

[[Imported from SVN: r10024]]

[[Imported from SVN: r8070]]

[[Imported from SVN: r8062]]

don't let the class inherit from gridfunction because there is no derivative type implemented for functions with values in manifolds

[[Imported from SVN: r7985]]

[[Imported from SVN: r7947]]

[[Imported from SVN: r7946]]

[[Imported from SVN: r7936]]

[[Imported from SVN: r7919]]

[[Imported from SVN: r7917]]

[[Imported from SVN: r7915]]

[[Imported from SVN: r7914]]