We should really clean up the mess that "reparamOnXXX" has become. Introducing one more horrible "haveParametrization()" member function to add to the general insanity :-)

ENABLE_SWIG has been replaced by ENABLE_WRAP_PYHTON and ENABLE_WRAP_JAVA.

- fixed MED node ordering
- polished 64 bit Cocoa Mac OS X FLTK version
- updated copyright

physical tags, partitioning, second order prisms

No commit message

No commit message

Changed the return type of GetOBB from SOrientedBoundingBox* to SOrientedBoundingBox

Emilie: added functions to remesh a volume

- fix bb computation to avoid model jumping around when adding points in z=0 plane
- pass interpol matrices as ref

- fixed missing getShapeFunction/getGradShapeFunction + return all sf
  at once so we can implement MElement in term of these consistently
  (w or w/o FunctionSpace)

- can now merge multiple meshes and remove duplicate mesh vertices
  (useful e.g. when merging several STLs)

- added easy interface to create model-based postpro datasets on the
  fly + use this for mesh quality stats

   1) this removes the asymmetry in all the routines involving mesh
      elements associated with points

   2) it makes it possible to use point-based datasets in the
      model-based post-processing module

  **********************************************************
    PLEASE TEST --- ESPECIALLY THE INPUT/OUTPUT ROUTINES!
  **********************************************************

* made GModel::getEntities() more efficient and generalized its use

CGNS I/O rewrite - alpha version

- removed every "throw;", i.e., every throw that did not throw an actual exception.
  throw wihout args should ONLY be used in a catch{} to re-throw the same exception
  (for it to be handled in another catch clause)

- introduced GmshMessage callback class so that clients can easily implement
  their own message handling (see example in driverOCC.cpp)

* implemented first rough draft of VTK mesh writer

introduce notion of entities in PViewData (in prepa for gmodel stuff)