copyright update

optimize ExportLc

test if there is anything to save before opening the file

generalized ExportLcField to all element types

prevent crash when no contour

- reworked the keyboard shortcuts with Laurent (our very own Macintosh
  Usability Guidelines Expert)

- new smoothing option for jpeg output

- new small GUI dialogs for ps/pdf/jpeg/gif/msh output (makes the
  'format' menu in the file dialog much shorter)

forgot to add a copy of the simpl pointer in the global simplex tree

- fix cast in a couple of error msg
- more fixes in elliptic algorithm (surf->TrsfVertices and surf->Vertices
  were incomplete)

fix+cleanup elliptic algorithm (didn't work when line mesh orientation was
opposite of original generatrice)

fix copy/paste bug in error message when quad/pyramids already exists

upgrade to Netgen 4.4

- moved the AdaptiveView stuff in its own header file

- fixed small bug in IsoSimplex (when Raise != 0)

- fixed small memory leak in Plugin/Extract (when one or more
  expressions are invalid)

This patch fixes an old design flaw in the post-processing module,
namely that we used a list of Post_View objects instead of a list of
pointers to Post_View objects in CTX.post.list.

This had many annoying consequences, in particular the fact that we
needed to be extra careful every time the list was reallocated (as
pointers to the list elements would become invalid).

I tried very hard to change the code everywhere it should be changed,
but I might have missed something. Please let me know if you see
anything suspicious (like a crash when you duplicate/remove/combine/...
post-processing views).

make Export_LcField also work with SimplexBase

The "Simplex" class is now derived from a new class "SimplexBase",
which contains only pointers to the nodes, and none of the other stuff
necessary for the actual mesh generation.

All the input/output/display routines have been rewritten in terms of
SimplexBase: this provides a gain of almost 50% in memory when you
just load meshes from .msh files to do visualization stuff, instead of
actually generating the mesh.

New "fast" routine to create simplices used only for
visualization. This speeds up the load time of large tetrahedral
(volume) meshes by roughly 50%, and brings Read_Mesh.cpp pretty close
to mshsort.cpp performance-wise. Since Read_Mesh does many more things
(all the Tree/List queries to create elementary/physical entites +
partitions), this is actually pretty good.

Loading a 1.4 million tets mesh file on my 1.5 GHz Linux machine now
takes about 17 seconds.

cleanup (and remove 2 shift/reduce conflicts in parser)

Fixed bug in second order mesh of closed curves

add error check in Oriente()

proof of concept for unproject (e.g. to add points with the mouse)

256 -> sizeof(string)

Don't call Mesh_Quality() in GetStatistics()

added vertex arrays in volumes to store the boundary of the cuts
when we draw them "as surfaces" (for Philou, not tested yet)

make depend

- fixed a couple of gcc warnings + added missing o->end() in
  jf's StructuralSolver plugin

- fixed the view option code: FLTK > 1.1.5rc1 breaks our use of
  widget->changed()...

- added background mesh support in the aniso algorithm

- added Orientation() members for hexas, prisms and pyramids

- use det3x3() instead of ad-hoc code everywhere it makes sense

make depend

- moved the smooth normals stuff into a separate file
- added smooth normal support for surface meshes (very nice!)

add an area constraint on the elements in the initial mesh to make
the refinement nicer

- Triangle now supports our background meshes

- fixed memory leak in Optimize_Netgen (forgot to free the volume
  vertices) + added missing 2nd order reset

- fixed a couple of typos in the docs

fixed old extrusion mesh generator (used the old 99999 volume hack!)

Removed the old test on ZonLayer: we actually WANT to use the automatic
volume if we set the layer number to 0. This is really pretty nice: we
can now get either automatic or manual numebring in all the extrusion
commands.

- All extrusion commands now return a list of 2 numbers (instead of 1):
  the first, as before, is the number of the "top" of the extruded region
  (i.e., a point for extrude point, a line for extrude line, ...), the
  second is the number of the "body" of the extruded region (i.e., a
  line for extrude point, a surface for extrude line, ...).

- "Extrude Surface" now always creates a new volume (automatically),
  EVEN WHEN THERE IS NO LAYERS SPECIFICATION. This makes it consistent
  with "Extrude Point" and "Extrude Line", which always create new
  curves and surfaces, respectively.

  Important Note: you will have to modify your old .geo files to avoid
  duplicate volume definitions if you use "Extrude Surface" without
  extruding the mesh (i.e., without the "Layers" command). These
  duplicate volumes would be harmless, but they would srew up your
  physical volume definitions later on...

  * Solution 1: use the new volumes (recommended). To do this, just
  remove your old extra volume definitions and let Gmsh create the
  extruded volumes for you. (To retrieve the volume number created by
  Gmsh, use "aa[] = Extrude Surface {...};;": the volume number is
  "aa[1]".)

  * Solution 2: keep the old volumes.

  a) clean way: retrieve the new volume number (aa[] = Extrude Surface
  {...};;) and delete the new volume with "Delete { Volume aa[1]; }"

  b) dirty (but handy) way: since, in order to create the new volumes
  with the less impact possible, Gmsh uses "low" numbers (actually,
  forcing "Geometry.OldNewreg=0") for the new volumes, just remove all
  "low number volumes". For example, if you have 4 "Extrude Surface" in
  your file, you can then just do "Delete{ Volume {1:4}; }"

Voila :-)

added sanity check for number of nodes per element

make depend + update docs

don't optimize netgen meshes by default + use Msg(FATAL) when something bad happens

added optimization hooks in the interface (GUI + comamnd line)

mesh optimization polish (redirect netgen messages to the GUI, etc.)