[[Imported from SVN: r7842]]

[[Imported from SVN: r7841]]

[[Imported from SVN: r7840]]

Computation of analytical energy gradients doesn't work, yet.

[[Imported from SVN: r7839]]

[[Imported from SVN: r7838]]

[[Imported from SVN: r7837]]

We need the extra flexibility for higher-order functions.
Run-time loss is around 1% -- that's okay.

[[Imported from SVN: r7836]]

[[Imported from SVN: r7620]]

[[Imported from SVN: r7617]]

[[Imported from SVN: r7616]]

[[Imported from SVN: r7614]]

[[Imported from SVN: r7613]]

[[Imported from SVN: r7611]]

[[Imported from SVN: r7576]]

notably the almost-equivalent UnitVector<4>: A vector with three entries
is interpreted by the UnitVector<4> class as a tangent vector given in
the basis returned by orthonormalFrame.  That is correct as it works
for all kinds of manifolds.  On the other hand, the Rotation<3> class
interprets as a 3x3 skew matrix (a tangent vector at the identity in
the matrix representation) even though the implementation used quaternion
coordinates.

To trick around this the use of exp methods with FieldVector<3> is avoided
by this patch.  A real fix would involve introducing a SkewMatrix class
and using that in Rotation<3>.

[[Imported from SVN: r7575]]

[[Imported from SVN: r7574]]

[[Imported from SVN: r7573]]

[[Imported from SVN: r7572]]

[[Imported from SVN: r7571]]

[[Imported from SVN: r7570]]

[[Imported from SVN: r7569]]

[[Imported from SVN: r7568]]

[[Imported from SVN: r7567]]

[[Imported from SVN: r7566]]

coordinates into a separate method, and fix an index bug on the way.
Now this derivative passes an fd check.

[[Imported from SVN: r7565]]

[[Imported from SVN: r7564]]

[[Imported from SVN: r7563]]

move method getFirstDerivativesOfDirectors into the rotation class.  It does not make sense for general quaternions

[[Imported from SVN: r7562]]

[[Imported from SVN: r7561]]

[[Imported from SVN: r7560]]

[[Imported from SVN: r7559]]

[[Imported from SVN: r7558]]

[[Imported from SVN: r7557]]

[[Imported from SVN: r7556]]

[[Imported from SVN: r7555]]

[[Imported from SVN: r7554]]

[[Imported from SVN: r7553]]

That space is a product space, and its LocalGeodesicFEFunction can be
implemented much more efficiently than the generic method.

[[Imported from SVN: r7551]]

and barycentric coordinates.  This is easy because we know the
number of element corners at compile time (I currently don't
see support for non-simplex elements at all).  This saves 
time, but not much: about 1%.

[[Imported from SVN: r7550]]

[[Imported from SVN: r7549]]