#717 Hyperelasticity/half-sum: make sure to zero current stiffness linear algebra before use (very important if for a given time iteration it is run more than once).

#714 FSI: modifying temporarily the code to deal with P1/P1. Currently numerical issue somewhere with nan values that pop up.

#713 Fork FSI to introduce a new one in which fluid and solid will be handled as two different meshes. So far the second mesh is the same as the first one and is not actually used, but it allows to check HappyHeart handles seemlessly another mesh without any conflict whatsoever in initialization.

#706 Geometry: even though the new static data attribute was one of the very rare case it was valid (there was no static initialization order fiasco as they were by construct built before the main due to the factory) I replaced them by a static method to be in line with the rest of the code.

#711 Improve slightly commit hook so that it's possible to emit a ticket from master if a ticket number is correctly specified.

#707 FSI: move in the policy the coefficient so that elasticity and hyperelasticity can work the same way. At the moment both model seems ok: the issue is due to the fact that in elasticity sources are added to the rhs and in hyperelasticity are removed from it.

#707 Hyperelasticity: remove the comment about the sign: it's clear it is correct (it matches hence the formula in Philippe's doc).

#707 Hyperelasticity: make sure residual is entirely computed within ComputeResidual (and likewise for tangent).

#707 Hyperelasticity: put VectorNewStiffness into the generic class rather than only for half sum scheme.

#707 ThirdParty/Petsc: fix a stupid bug in an assert (program-wise size rather than processor-wise one was checked).

#707 Hyperelasticity: move into ComputeResidual() and ComputeTangent() the applciation of boundary conditions.

#707 Petsc: I removed the Swap with a raw Vector, as I was not completely sure whether it was really safe. I put instead a method to fill content of a given Vector from a raw Vec.

#707 Elasticity/hyperelasticity: play with Lua files to compare them and also compare them to Frefem.

#706 Remove the transformation GeomElt -> RefGeomElt from GeometricEltFactory, as each GeomElt now provides a direct link.

#704 RefFEltSpace: add a more convenient accessor that fetch the correct ExtendedUnknown given an Unknown and a finite element space.

#704 Hark back on ExtendedQuadraturePoint, which was a dreadful idea: the additional content besides the quadrature point were specific to a given LocalVariationalOperator, and thus had no business begin highlighted at a more generic level.