drawing code from another widget toolkit)

Small update on reduceComplex and coreduceComplex methods to correspond current
insight of reduction and coreduction: coreduction is the same thing to the dual
cell complex as reduction is to the primal cell complex, and hence has no
"priviledges" over that.

When coreduction is applied to primal complex, we obtain the thick cuts
of generators of homology groups obtained via reduction of the primal
complex.

Emilie: added functions to remesh a volume

Classification editor detects all closed or open curves

  the .msh format -> bumped version from 2.0 to 2.1)
- move createTopologyFromMSH into GModel.cpp (and renamed to
  createTopologyFromMesh)

Added cell combining methods (about O(n^2)) to further reduce the size of
the cell complex before applying Smith normal form computation.

(Don't know why there seems to be some benchmarks/ files modified too,
Somebody else committed between my update and commit?)

added createtopologyFromMSH to create Topology from list of triangles

Fixed error in line curvatures

about O(n^2) -> O(n*log(n)).

The bottleneck is now in the Smith normal form computation of the reduced
cell complexes, O(n^3).

- uniformize syntax for parFromPoint/XYZtoU (cf. versions in GFace)
- fix compile

Added ability to compute the "scalar potential cuts" for homology
generators using coreduction. Modified utils/misc/celldriver.cpp to
demonstrate this.

Added some more or less useful utility and experimental functions.
Numerious bug fixes.

added createTopology when Merging a MSH File and implemented discreteEdge