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

[[Imported from SVN: r7839]]

[[Imported from SVN: r7837]]

[[Imported from SVN: r7617]]

[[Imported from SVN: r7611]]

[[Imported from SVN: r7553]]

[[Imported from SVN: r7416]]

using a finite-difference approximation only for the derivative
of the gfe gradients wrt coefficients.

[[Imported from SVN: r7230]]

[[Imported from SVN: r7228]]

[[Imported from SVN: r7227]]

[[Imported from SVN: r7131]]

[[Imported from SVN: r7030]]

[[Imported from SVN: r7024]]

transpose one gradient matrix.  This makes the code run, but it still doesn't work even for RealTuple<1>

[[Imported from SVN: r6547]]

[[Imported from SVN: r6500]]

[[Imported from SVN: r6498]]

übersetzt jetzt und stürzt nicht ab.  Es kommt allerdings noch Unfug raus.

[[Imported from SVN: r6451]]

[[Imported from SVN: r6297]]

[[Imported from SVN: r5931]]

[[Imported from SVN: r5807]]

[[Imported from SVN: r5741]]

[[Imported from SVN: r5594]]

[[Imported from SVN: r5567]]

[[Imported from SVN: r5566]]

[[Imported from SVN: r4472]]

[[Imported from SVN: r4077]]

[[Imported from SVN: r4069]]

[[Imported from SVN: r4059]]

[[Imported from SVN: r4034]]

[[Imported from SVN: r4033]]

[[Imported from SVN: r4023]]

[[Imported from SVN: r4022]]

[[Imported from SVN: r4021]]

[[Imported from SVN: r4020]]

now works on a per-element basis and has been moved to the local rod assembler.
That way, it has lost most of its 1d-ness.  In the future, it will be usable
for other problems as well.

[[Imported from SVN: r4019]]

[[Imported from SVN: r4016]]

The class 'Configuration' is now called 'RigidBodyMotion', and can be parametrized by the space dimension

[[Imported from SVN: r4003]]

[[Imported from SVN: r4001]]