version 3.47 (1/june/2016) 
 - tools to build new FiniteElement2d and FiniteElement3d from quadrature formula
    
     FiniteElement2d EFQF(qf23pT);
     fespace Zh(Th2d,EFQF);

     FiniteElement3d EFQF3(qfVp14);
     fespace Zh(Th3d,EFQF3);

 - add solveurs FETI and  BDD in  HPDDM
 - add savegmsh in gmsh plug in to write file for gmsh (thank to Lo Sala salalo80@gmail.com)
  see  ttestio.edp for the examples 
version 3.46  
version 3.45-1 
 - add new finite element to store value a quadradure point.
   in plugin  Element_QF  (see exemples++load/Element_QF.edp) 
   in 2d FEQF1,FEQF2,FEQF5,FEQF7,FEQF9  corresponding resp. to  quadrature  
         qf1pT,qf2pT,qf5pT,qf7pT,qf9pT
   in 3d FEQF13d,FEQF23d,FEQF53d  corresponding resp. to  quadrature  
         qfV1   ,qfV2  ,  qfV5
   and FEQF and FEQF3d for the default quadrature. 
 - ajout gluemesh of 3d mesh 
 - ajout operator   b= A* c; // where is real matrix avec b and c complexe vector. 
version 3.45 (10 march 2016)
 - correct veru old bug in ffglut, some time we miss wait due to asyncronuos stuff 
    and we can hav e memory error for the same reason .
 - try use lot of preinstall software  (hard job for ubuntu and debian)
 - pass to tetgen version 1.5.1-beta (the last one) 
 - add Th.hmin and Th.hmax methode to Mesh type 
version 3.44-1
 - add new hpddm example for Stokes probleme (Thank to P. Jolivet )
 - put the new convect 3d as the default, (newconvect do nothing now). 25/02/16
 - try to add new version do 3d convect ( add newconvect=1;) in test ...
 
version 3.44 (22 feb. 2016) 
 - add tools to extract border after movemesh in Curvature  plugin)
     see curvature.edp, can be useful to remesh after movemesh. 
 - restruct the hpddm and petsc plugin ( thanks to P. Jolivet)
   add lots hpddm and petsc examples in examples++-hpddm
version 3.43-33 
  - remove some spurious print when verbosity = 0 
  - add initialisation of array which string parameter like in
     string[string] b= ["-1",1,"13","qsdqdq"];
     int[string] a=["2",1,"12",4];
  - add setw iomap stuff like in:
     cout << setw(10) << 1 << " " << 3;

version 3.34-2
  - correct examples++-load/ff-get-dep.awk in case of choose of lib in syntaxe of library dependes in 
    WHERE-LIBRARY stuff 
  like in : 
  //ff-c++-LIBRARY-dep: cxx11  [petsc|mumps parmetis  ptscotch scotch]  scalapack blas [mkl]   mpifc  fc mpi  pthread
version 3.43-1 
  - correct PB with hpddm (pass to git version d2a99d89ee37777d2423da218b29e2080950da1e Jan 20 22:03:12 2016)
  - correct PB of FLIBS
  - pass to evrsion 3.43-1  3 feb 2015 
version 3.43 (version 3.43 (3626:8e3d8953975d), 2 feb 2016 16;00 CET))
  - change  RT03d finite element interpolation (put 1 point integration by face)
  - add renum=false in all trunc fonction of DDM examples 
  - warning change the defaut no renumbering to be sure trunc commute 
       the two mesh must be the same  truc(trunc(Th,I),J) == truc(trunc(Th,J),I) 
remove this version (version 3.43 (3623:882bfbae1bf6, feb 2016 15:30 CET))
  - change the trunc function in 2d to keep the ordre of old vertices to remove signe probleme
    in some restriction operator of RT finite element  for example
  - correct of the formal operator trace (bug was comute  a00 + trace(A), thank to O. Pironneau )
  - put hppdm source include in the distrubition to have more safe compile process.
  - update a little the documentation (add new finite element)
  - add new finite elemnt just for interpolate partition of unity in DDM / HPDDM method 
     Edge23ds0, Edge13ds0,Edge03d in plugin  Element_Mixte3d
  - add new finite element P1bl, P1bl3d in plugin  Element_P1bl
  - add new finite element Edge23d, P1bl,in plugin  Element_Mixte3d
  - change EPSD in eigenv because touble in some very rare case (see G. Vergez)
  - correct problem for vectoral solution on implicity declaration of product of simple  EF like  P1b
  - add tools to compute curvature of border in Cuvature plugin  
  - all implicity loop  for lot of array  : real[ini], real[int,int], real[string] ,matrix , string[string]
  the syntaxe is 
  if B is like a  Array : 
  for [i,bi: A] bi=f(i);

  if A is like a matrix : 
  for [i,j,aij: A] aij=g(i,j);
  see forall.edp example file for more exemple 
  - change the chinise documentation (thank to Helin GONG) 
version 3.42 
 - correct some Makefile 
 - correct optimisation of convect in case on windows bug)
 - add tool do build quadrature formule when we split triangle in 3 triangle for HCT element.
 - add new quadrature formula 3d exact form degre 6 to 14 in qf11to25.cpp 
   the name of  qfV quadrature are  qfVp1 .. qfVp14 
 - correct optimisation in convect for navier stokes problem..
  for this kind of expression
     + int2d(Th) ( -alpha*convect([up1,up2],-dt,up1)*v1 -alpha*convect([up1,up2],-dt,up2)*v2 )
   in this case now only on characteristics is computed 
 - correct trouble in configure this /usr/bin/machine 
 - map&key to know if a key exist in a map int[string], ....  
version 3.41
 - add tools to get mesh of fespace Vh in 2d and 3d 
     mesh th2= Vh2d.Th ;  mesh3 th3= Vh3d.Th ; 
     not the type of the mesh is a const pointeur   => lot of cosmetic change
version 3.40-2
 - try to resolve problem with petsc metis and freefem++ metis
     now freefem++ use petsc lib if petsc is find for
       parmetis, metis,ptscotch,scotch,fftw3_mpi, mumps,scalapack
 - correct hpddm examples
 - rename plugin schwarz in hpddm 
 - correct problem on compilation do scotch 
version 3.40-1 
 -  move of all hpddm code un examples  examples++-hpddm ( Own by  P. Jolivet and F. Hecht)
version 3.40 (14 sep. 2015)
 - add new C1 Finite element HCT see examples++-load/bilabHCT.edp 
 - change permission in install example++ for read and write under windows 
 - try to build a windows 64 version under msys64 system ...
 - Add use ArrowSize= of size
version 3.39
 - remove compile flags -fPIC under windows 
 - pas to version MUMPS-5.0.1
 - correct some problem of memory leak, now the decrement method in RefCount call delete
    and all decrement are change in   Add2StackOfPtr2FreeRC(stack,m); (more safe). 
    so the previous plugin become incompatible. 
version 3.38-2
 - add hppdm explain in the doc. 
 - add function cube to simplify the construction of mesh of cube like square in 2d 
      see cube.edp example. 
 - add function ClosePoint in ClosePoint plugin
 - add += example of varf   
 - correct mortar example: replace P0Egde in P0 => better result  
 - correct map func if dfft plugin 
 - add splitComm for hpddm
version 3.38-1 
 -  add finite element Edge13d (Finite Element of degrees 1) Thanks to marcella@bonazzoli.it
     and exemples of wave guide in waveguide.edp
 - correct renumbering function in case of rhs 
 - correct new type of optimisation in integral, (version 3.38 25june 2015)
  optimized=2 => do optimisation without check 
  can be useful in case of random problem. 
 - correct convect operator in 3d (remove randow part, 
   to remove problem of optimisation in varf.
   
version 3.38
 - add parameter in renumbering function. 
version 3.37-1 (22 mai 2015)
 - clean examples-mpi, remove of all usage to RemoveDOF,   bb2d bb3d, findDiff   (see version 3.36)
 - update schwarz.cpp for new hpddm version 
 - add read of real[int,int], int[int,int], complex[int,int] .. 
version 30 april 2015 (3319:8657e0526391)
 - correct freefem++ launch  in windows (remove wait when launch vai freefem++)
   add wait in case of launch through launchff++.exe 
 - in DG linear form with jump or mean in test fonction was wrong 
   like varf a(u,v)=intalledges(Th)(jump(v)*u) ;

 - correct problem of try/catch in freefem++ func 
version 3.36-2
 - correct problem some  null ref  with compile Apple LLVM version 6.1.0 (clang-602.0.49) (based on LLVM 3.6.0svn)   xcode 6.3 
   some warning exist but all test pass. 
 - correct problem of cleaning tmp memory for small array 

 - correct configure and makefile for windows compilation process
version 3.36 
 - add tools to remove  (now build in)
   utility.dylib removeDOF.dylib  symmetrizeCSR.dylib 
   bb2d bb3d -> boundingbox
   fast -> ltab=lround(dtab)
   Unique to get all unique value of a array (findDiff)   
   
 - after version 3.33 the some compilation flags a lose, correct configure.ac
    Please do not use version from [3.33 .. 3.35] included. 



 - add disable-gmm configure flags 
version 3.35 dist (12/3/15) (rev 3246:664a6473d705) (warning slow version)
 - optional lib seach  [toto|tyty] in WHERE-LIBRARY seach lib 
 - change the metis to scotch-metis interface in hips and parms .. 
 - change the current version of scotch metis mumps parmetis 
 - remark: .dll , .so, .dylib are incompatible with previous version.
 - ff-run.in under windows (use src/bin-win32/FreeFem++-mpi 
   and not src/mpi/FreeFem++-mpi for dll problem). 
 - correction init plugin missing static in int DoLoadInit() ; 
version 3.34-2 (20/2/2015) (warning slow version)
 - do correction do compilation for mpi version under window
     big change in ffapi.cpp rules ( use now call back) end init in mymain. 
     schwarz plugin run under windows without petsc  lob. 
 - add some function to then  gsl interface  (airy,  biry, ...)
version 3.34-1 
 - correct lot of mistake for simple compilation of 
    hpddm interface ..
 - add no mandatory lib for petsc 
      write the WHERE-LIBRARY search lib in awk (more simple ) 
 version 3.34 (warning slow version)
 - configure : 6/02/ 2015
    correct mpi for  sgi uv computer
    find gsl lib 
    add find petsc (in progress) 
    add missing file in distribution
 -  correct for compilation with g++-4.9.1 -std=c++11 ( without downlaod) 

BUG in configure => slow version after until version 3.36 ,  25 mars 2015 ... (dur Dur)





  no optimisation flags in compilation

 - add hd5 interface  (13/01/2015) 
    Thank to Mathieu Cloirec CINES - http://www.cines.fr
   voir example iohd5-beam-2d.edp	    iohd5-beam-3d.edp
 - add find of libgsl  in configure script 
 - correct pb of memory leak in case 
   matrix A = ...; in loop ( this occur in lot of case, tanks to P. Jolivet )
   correct small memory leak in use of routine due to debugstack. 
 - correct bug in periodic condition in case common dof with periodic.  
 - correct big bug in memory management of sparse matrix

 version 3.32 
 - correct of problem of plugin and mpi,
    build all dynamics lib with and without mpi, 
    the mpi version is install  dir lib/mpi 
 - correct of plugin MUMPS.cpp for complex value. 
 - add vectorial operator a/v  and v/a where a is scalar and v vector like real[int], ...   

 version 3.31-3 (rev 3053:4e164226411d 12 Aug 2014) 
 - correct the problem of size of arrow in 2d plot 
  version  3.31-2 (rev 3052, 11 July 2014) 
 - correct stop test function in LinearGC (for zuqi.tang@inria.fr)
  build tar.gz distribution (rev 3050) 
  build version MacOs 3.31-1
 - correct bug put in DG formulation (rev 3044)
   jump, mean , was wrong from Sun Jun 29 22:39:20 2014 +0200 rev 3028
  version  3.31-1 (rev 3042, 10 july 2014) 
 - function to put your stop test in LinearGC and NLGC
   the prototype is 
    func  bool stop(int iter,real[int] u,real[int] g)	;
    	  {    return g.linfty < 1e-5 || iter > 15;}
    LinearCG(DJ,u,eps=1.e-15,nbiter=20,precon=matId,verbosity=50,stop=stop);
    

 - add functionnal  interface  to arpack (Eigen Value) 

    func real[int] FA(real[int] & u) { real[int] Au=A^-1*u;return Au;}
    func real[int] FB(real[int] & u) { real[int] Au=B*u;return Au;}
    int k=EigenValue(n,FA,FB,sym=true,sigma=sigma,value=ev,vector=eV,tol=1e-10,maxit=0,ncv=0);

  see examples++-eigen/LapEigenValueFunc.edp for a true example


version 3.31  (rev 3037, 1 july 2014)
 - re-add tan function for complex number  
 - correct a big mistake in LinearGMRES , the result are completely wrong,
   correct also the algo.edp
 -  add sqr functon of O. Pironneau
 - correct update of mercurial depot (rev 3034, 1 july 2014)
 - correct misstake  in examples++-3d/MeshSurface.idp 
   about   computation metric in function build surface  Sphere mesh
   and add a function to build Ellipsoide surface mesh 
  func mesh3 Ellipsoide (real RX,real RY, real RZ,real h,int L,int orientation)
 - correct a bug the DG with periodic boundary condition with only one 
   layer of element.  
 - add pluging "bfstream" to write and read in binary file (long, double, complex<double> and array)
   see bfstream.edp for an example. 

version 3.30-1 may/2014 ( hg rev: 3017)
 - add levelset integral on 3d case ( on levelset and under level set) 
 - correct problem with Ipopt / lapack configure ...
 - add BEC plugin of Bose-Enstein Optimisation
 - standardisation movemesh3 -> movemesh ( same parameter of 2d version )	
 - correct jump in basic integrale to be compatible with varf definition
    jump is not zero on boundary, it is - boundary value ..
    fespace Ph(Th,P0);u= 1;int1d(Th)( jump(u) ) = - length border . 
 
30/6/14 (ALH)
 - [[file:download/getall]]
   - now checks downloaded packages against known MD5 values
   - new -h inline help option

version 3.30. ( for windows hg  rev : 3013)
 - add binary ios:mode constant, to open file in binary mode under window
    to solve pb of seekg under windows  
 - add multi border april 23 2014 , (hg rev : 3004)
    syntaxe example:  
  // i is the index of the the multi index 
  // the number of sub border is given by teh size if the array to set the number of seg on the border ..
  // so for multi border the  number of seg of the border , must be  in int[int] array 
  real[int] RC=[ 0.1, 0.05, 0.05, 0.1],
            XC= [0.2,0.8,0.2,0.8],
            YC= [0.2,0.8,0.8,0.2];
  int[int]  NC=[-10,-11,-12,13]; 

  border bb(t=0,1;i) 
  {
  //  cout << " i = " << i << endl; 
    int ii = (i+1)%4; real t1 = 1-t;
    x = xx[i]*t1 + xx[ii]*t;
    y =  yy[i]*t1 + yy[ii]*t;
    label = 0; ; 
  }
  border cc(t=0,2*pi;i) 
{
	x = RC[i]*cos(t)+XC[i];
	y = RC[i]*sin(t)+YC[i];
	label = i+1; 
}
  int[int] nn=[4,4,5,7];
  plot(bb(nn),cc(NC),wait=1);
  mesh th= buildmesh(bb(nn)+cc(NC)) ; 
  plot(th,wait=1); 
  
 - add ltime() (rev 2982)  function returns the value of time in seconds 
       since 0 hours, 0 minutes, 0 seconds, January 1, 1970, (int) 
 - add new macro tool  like in C  (rev 2980)
    FILE,LINE,Stringification() to get line number and edp filename,
    add quote to a parameter after macro generation … 
    Like in 
    cout << "in " << FILE << " line " << LINE << "    -- '" 
         << Stringification( "zzz" aa  () {} +  /* */  bb cc) << "'" << endl;
  add new int2d on levelset  in 3d  (int test)
  add basic  func mesh3 Cube(int nx,int ny,int nz) in cube.idp file.
version 3.29  (hg rev 2973)
  - add int storagetotal(); and int storageused(); function 
     to get static of malloc ( memory storage usage)
   
  - correct problem of region evalution in jump and mean function 
version 3.28  ( merge with freefem++-cs tool)
  - add download/getall perl script to download all related soft
  - add  int1d on isoline for matrix ...
  
version 3.27 
 - correct bug in display of P1b finite element in 3d 
   error in SplitMesh<R3>  function  (Thank to O. Pironneau) 
 - add  AddLayers(Th,suppi[],sizeoverlaps,unssd[]);
 - add  tool to trunc to get element numbering for Thn to Tho 
    int[int] In2o(1),Io2n(1); 
    Tho =  trunc(Tho, x<0, new2old=In2o, old2new=Io2n);
 - add restrict function for get dof numbering old to new
    fespace Vnh(Thn,Pk), Voh(Tho,Pk);
    int[int] n2o=restrict(Vnh,Voh,n2ok);     
 - correct mistake in gsl interface random number (9/1/14)
     missing all random  distribution ..
 - add interface  with gsl random number generation in test .. 
 - correct pb of compilation under window (mingw32/msys) 
6/2/14 (ALH)
 - Creating documentation hyperlinks (in emacs orgtime and in Doxygen format)
 - New script build/orgindex and file index.org list all hyperlinks in the FF source in Emacs org-mode format
 - Set all hyperlinks to relative paths to make sure that they work from any location
 - Separate download script for all third-party software [[file:download/getall]] (request from FH)
 - created common makefile goals in [[file:download/common.mak]] for all downloaded packages (request from FH)
   - changed [[file:download/arpack/Makefile.am]] to make use of [[file:download/common.mak]]
   - done for [[file:download/blas/Makefile.am]] as well
 - added option --enable-mkl-mlt in [[file:configure.ac::enable_mkl_mlt]] to allow FF to be linked with the
   multithreaded MKL when an external library requires it (request from Atsushi Suzuki)
version 3.26-3  09/12/2013 
   correct problem with openblas need pthread lib 
01/12/13
    install dowload lib 
    update lapack interface in examples.
    add WHERE_LIBRARY-file in install 
    change the order of seach lib with   WHERE_LIBRARY-download 
       WHERE_LIBRARY-config        WHERE_LIBRARY
    remove all relative in WHERE_LIBRARY-config
    correct of superlu compile problem
13/11/203  Pass to version 3.26 ...
  - correct pb of ambiguity with new complex lib 
31/10/2013  compilation on MacOS 10.9 with compile of xcode 5.0.1 
  - correct the pipe.cpp to compile with clang-500.2.79 
  - remove of warning message genrate by clang++ -std=c++11 (version clang-500.2.79)
  - correct also compilation problem 
  - correct error in case of block matrix with block of 0 row or 0 column. 
21/10/2013 (ALH)
 - umfpack configuration cleanup (request from Fred)
 - blas configuration cleanup (request from Cico)
   - configuration option --enable-generic-blas has no effect anymore, removed
   - Atlas compilation directives have no effect anymore, removed
   - Moved OpenBLAS compilation from FFCS source tree to FF source tree
 - Added configure option --disable-system-fftw to skip any default FFTW library installed on the system
 - Changed location of lapack WHERE-library setup in [[file:configure.ac::WHERE_lapack]] to insure that it is always
   defined (request from Fred)
 - changed FFCS Fortran MPI configuration for MPICH2
 - split DOC makefiles into Makefile.am and figs.mak to suppress automake warnings about portability
 - moved build/download script from FFCS to FF before reorganizing FF downloads (request from FH)
10/9/2013 (ALH)
 - Corrected pastix compilation for FFCS on MacOS 10.6
 - Removed dependency from lg.ypp to lg.tab.?pp in [[file:src/lglib/Makefile.am::lg.tab.?pp]] to avoid automatically
   generated conflicts
05/09/2013 put version .tar.gz on the web 
  - correct makefile wget in pastis and superludist  
  - correct compile of load example :  paradiso, gsl automaticaly 
 - add AutoGeneratedFile.tar.gz  a file contening all 
    file build by autoreconf in case of non automake tools (1.13)
 - add missing file in dist or in mercurial distribution
 - correct download of scotch un curl 
 - correct problem in a*[b,c, ... ]' in case of complex value   
5/9/2013 (ALH)
 - force Umfpack build to run sequentially (parallel make crashes with "pipe from processes is a directory"?)
 - pARMS download URL changed to http://www-users.cs.umn.edu/~saad/software/pARMS/pARMS_2.2.php (thanks Fred)
 - corrected download/mumps parallel compilation (WHERE was not built properly)
 - Removed all Mercurial-tracked files from .hgignore
 - added options --with-[package]-include= and --with-[package]-ldflags= to avoid downloading existing packages
 - Parallelized MUMPS compilation in download/mumps (and mumps-seq)
 - Applied a patch from Fred for compiling SuiteSparse on Slackware64-14.0
 - Added configuration option --enable-hypre (disabled by default, contrary to other tools)
 - Deactivate FFTW download when a local version is found (request from Helmut Jarausch)
10/7/2013  (version 3.25)
 - remove of Makefile.in configure for the hg distrubion 
    use :  autoreconf -i  # too build Makefile.in 
         before to configure 
         need automake version 1.13 ... 
 - merge FFCS (ALH) version and  ff++ version (FH) of freefem++ programs
 - remove all automake file form the hg data base
 
 - add new parameter to ffglut for demo of freefem++ 
    ffglut  [-nv|-v|-vv|-vvv] [-wait 0.5] [-g 512x300+10+10] [-t title] [file]
    all number can be change the wait is in second 0 is default value nowait bewteew plot
    -g 512x300+10+10 is the geometry of the  graphic window
    -t the title of the windows 
26/6/2013 (ALH)
 - created a build/ subdirectory for build tools
 - enabled parallel make ("make -j")
 - created a separate file (acmpi.m4) for complex MPI configuration options
 - added configuration option --enable-ffcs to make the FF source compatible with FFCS
 - backported all current FreeFem++ patches for FreeFem++-cs into the FreeFem++
 - started main doxygen page mainpage.dox
9/06/2013
 - correct extract function of mesh Lo Sala <salalo80@gmail.com>
 - correct int2d on levelset  see example intlevelset.edp 
 - correct automake TESTING part (in progress) 
 - correct typo on the doc with .*= ./=  operator
 - correct bug in RT0 3d , code in the construction of the DOF.
   the bug is all dof on border  face of same element have same dof number.
   thank to Laurent Bernard <laurent.bernard@lgep.supelec.fr> 
version 3.23
 - do cleanning in version remove x11, glx, std : freefem++
   clean compile option for clang and clang++ compiler 

 - add flags to remove internal boundary in 2d,3d in function change 
     rmInternalEdges=1
 - correct glumesh in case of no mesh in 2d
version 3.22
 - add multi windows graphics ; WindowIndex=0 in plot function
   add new event in graphic windows 
     * to set/unset  default graphics stat to previous plot 
 - add getenv, setenv , unsetenv function in shell plugins for gestion of environnemnt variable for 
   openmp. 
 - correct pb un trunc for 3d mesh with too flat element (sliver) , and cleaning code .
 - correct bug in gestion off outsite flag is 3d in case of brute force (searchMethod>0) 
version 3.21-1 
 - correct bug  a=b' ; of full matrix 
 - bug in assert on 3d mesh of huge dimension with periodic condition.
 - correct build Delaunay 3d mesh of set of point. 
   see examples++-load/convexehull3d.edp
version 3.21 frev 2013
 - n, resize in array of finite element function ... 
     see edp  array.edp file 
 - correct pb of compilation of Ipopt with clang++ -std=c++11
 - correct une NSCahouetChabart.epd examples , and correct un doc. 
    this is a complete rewritting. FH. 12/02/2013.
 - correct in change function the "flabel=" parameter in 2d and 3d 
version 3.20-3
  - add master= in solver of MUMPS interface to set the master mpirank 
        <0 => distributed matrix. 
  - correct problem in label generation in freeyams (18/01/2013)
     all label border was set on 1  before. 
version 3.20-2
  - add MUMPS parallel version (in test)
  - add paradiso seq solver ..
version 3.20
 - correct isoline plug in case of sadlle point
 - change the compilation tools under windows gcc 4/7 + freeglut / ...)
 - change the compile tools on mac pass to clang++
 - compile ok of c++11  compiler
 - add formal tools on array [] or matrix [[],[],]  for elastic problem. 
   let A a formal array 
   if A is matrix 2x2 or 3x3 : trace(A) , det(A) ,    Cofactor(A) ; 
    A:A  = sum_ij A_ij * A_ij 
    2*A,   A*2 // muliplication by a scalar 

 -add integration on levelset line (in test) 
 
version 3.19-2
 - correct pb of C in/output in pluging (in test) 
 - correct bugs  mshmet pluging in case of double Eigen value 
 - correct typo problem (string size) when a change the default
    window size in postscript in version 3.19-1 to have more 
    precise postscript plot
version 3.19-1
 - add tool to create Quadrature formular 1d,2d,3d with 
   plugind: load "qf11to25"
   real[int,int] tab1(np,2),tab2(np,3),tab3(np,4);
   QF1 qfe1(norder,tab1);// 1d
   QF2 qfe1(norder,tab2);// 2D
   QF3 qfe1(norder,tab3);// 3D
   where tab(0,i) = weigth , tab(j,i) =  j coord, norder  order of the
     quadrature 
   see examples++-load/LaplaceP4.edp for example 
 - correct download auto compile of
    mmg3d, mshmet, scotch, ...
version 3.19 (20 april 2012) 
 - scotch partitionner interface see scotch.edp in examples++-load
 - add isNaN(x),  isInf(x), IsNormal(x)  function to check floating point
      of real value x, see ISO C99. 
 - add function : NaN() and  NaN("") to build NaN real number (double in C) . 
 - correct error in macro with operator ./=  and .*= 
 - add Ipopt doc (thanks to Sylvain  Auliac)
 - add Ipopt interface (thanks to Sylvain  Auliac)
 - correct 3d trunc bug in case of internal boundary
      thank to Yoshihiro Tomita <tomita.yoshihiro@gmail.com>. 
  - add new type of array , array of array
    see taboftab:edp in  examples++-tuturial
    real[int] a;  
    real[int,int][int] m(10);
    real[int][int] v(10);
   not well tested.
version 3.18-2
 - add plugins with sequentiel  mumps without mpi 
 - add conversion of re and im part of complex sparse matrix
   A = C.im; 
   A = C.re; 
version 3.18-1
 - correct Typo error in example 
 - add generation of error in case of periodic boundary condition 
   non scalar problem. 
 - add tools for adaptation of P2 and P3 finite elements with metrics
   see APk-AdaptEpsDeltaPk.edp APk-FreeFemQA.edp	 APk-MetricPk.edp
    APk-ExplicitPkTest.edp	       APk-LaplaceDirac.edp

 - New example in in Chapter 3 Navier Stokes Newton NSNewton.edp
 - add cod to build matrix and vector form varf in 3d in case of
      different meshes (in test )
 - correct NSprojection.edp chap3 example (PB in out flow BC.) 
 - correct compile of mmg3d v4 plugins (small change un distrib archive) 
v 3.18 (11/01/2012)
 - rewrite the isoline-P1 plugins (new name isoline.{dll,so,dylib} )
     see example Leman-mesh.edp and isoline.edp in 
     examples++-load directory 
 - correct bug in cas of resize of array with 2 index (full matrix)
 - correct assert in MPI  in gather and scatter
 - correct bug in case of return  in loop for or while. 
 - correct a=int2d(Th,l)(1) in case simple expression  when l is a  array.
 - build a pkg under macos for distribution .
v 3.17  (17/11/2011) 
 - correct PB of pugins: MUMPS, parmetis, metis, mmg34, mshmet 
 - the new load interface  (for more safe IO)
 - build mpi for windows with msmpi ( begin) 
    memory error after end .. 
 - remove add by default $LIB_MPI in ff-c++ 
 - change the way add thing in freefem++ in case of  dynamics load 
   to be  compatible   plugin with freefem++cs  ( solve pb with
     cin,cout, cerr in  dll may be..).
   LOADINIT(Init); // where init is a class 
or do 
   addingInitFunct FFinit(100,ffinit,"MUMPS_FreeFem"); // where ffinit the init function ...
v 3.16-1
 - cmaes interface in scalar and MPI case  (thank to S. Auliac)
    see doc and examples : cmaes-mpi-VarIneq.edp , cmaes--VarIneq.edp
 - add NLopt interface (thank to S. Auliac)
v 3.16 
  technical stuff: 
  - correct auto load of mumps
  - add tool to do automatic load or static load (for testing)
    see MUMPS_*cpp example 
v 3.15 
  - correct all examples++  load and 3d rebuilde the all.edp 
    and test this examples. 
  - correct version old bug when full matrix (arry)
    A=A;' is now correct (set and  initialization )
    and add  A+=A';    A-=A';

  - reput metis 4.0 form netlib ... (pb of compatibilite with other // soft hips)
v 3.14-1
  -   change interface with metis 5.0.1 
  -   Complete writing of mmg3d interface with version 4 of mmg3d
      le plugin is "mmg3d-v4.0", the parametre are the same have 
      command line mmg3d , execpt the file name are not given. 
  -   remove old bug 3D in interpolation P1 operator 
      correct bug 29/08/2011 (thanks to rychet@fzu.cz)
      remove wrong bulid of KHat (memory out of bound)

v 3.14 
  - correct in configure remove the default -O2 -g autoconf value 
    and add -g in case of --enable-debug 
  - a very old error in the on() functionnal
    the bug is with  a vectorial finite element like RT0, .. 
     the name of vectorial must     be in lexicographic order 
     so   u1,u2  is ok but v,u is wrong
    is correct in version 3.14 the 24/08/2011

  - correct trap in check convect-apt.edp  example
  - add the existance of patch in configure 
v 3.13-3 ( 30 june 2011 Seville)
  - correct the Hips interface (not to bad , in test)
     load "hips_FreeFem"
     int[int] iparm(1);real[int] dparm(1);
     HipsDefaults(iparm,dparm);) // set def option 
     ( limit to 100 Hips active linear system ).
  - add interface with MUMPS_4.10.0 version  (with automatic download )
  - add P1dc3d finite element in pluging  "Element_P1dc1" 
  - correct bug in gibbs renumbering in 1 case veru sample mesh)
  - correct bug in interpolation operator with periodic BC
    mesh Th1=square(2,1), Th2=square(2,2);
    fespace Vh2(Th2, P1),Ph2(Th&,P1,periodic=[[1,x],[3,x]]);//  une couche
    matrix Jh=interpolate(Vh2,Ph2,op=0,inside=0);
    Ph2 w=1.;Vh2 wi; wi[]= Jh*w; // wi must be 1, now this ok. 
v 3.13-1
   - correct compilation problem on fedora 13 WITH MPI
v 3.13 (25 may 2011)   
   - update the finite element list in  documentation 
   - add (load "Element-Mixte")  NEW FINITE ELEMENT 2D
         TDNSS1    sym matrix 2x2 conforme in $\{H(div div) / div(div s)) \in H^{-1} \}$
         RT1 and BDM1  conforme in H(div) Raviert Thomas of Degre 1 and Bezzi, Douglas, Marini 
         RT1 and BDM1ortho   conforme in H(curl) 2 Nedelec Finite Element
v 3.12-3 
   - new finite element in 2d for elasticity in test. TD-NSS0 (see these of Astrid Sabine Sinwel)
      A New Family of Mixed Finite Elements for Elasticity
   
   - add matrice matrix multiply in lapack interface 
   - add tool to change the region and label number 
       change(Th,fregion= integer function to set the region number )
   - nuTriangle now given the tet number in 3D.  
   - add mpireduce of matrix to build parallel matrix
v 3.12-1 april/2011 
      see remove examples++-mpi/chaleur3D-superludist.edp
   - correct the precond of gmres algo in A^-1 : (29 mars 2011) to build a correct Navier-Stokes 
   - add cast TypeOfSolver  on int to parametrize the choice of linear solver.
   - correct intersection of given metrix in adaptmesh (bug introduc in 07/10 verion 3.9 
     in case of convect-apt  example  
v 3.12-1 10 10 fevr 2011 
    - add VTK write for paraview, examples++-load/VTK_writer.cpp		examples++-load/VTK_writer_3d.cpp (see source for moe details) 
    - put comment in the documentation about negative tgv 
    - correct pb blacs mkl  (unde with mumps) 
    - correct mpiReduce for complex and real data.  
    - add mpiAllReduce(umax,dmaxg,comm,mpiMAX);  where for real umax,dmaxg;
    - add tag verson 3.12-win32 
    - add inferface for mpi in win32 architecture (form 32 version)
v 3.12    17 jan 2011
    - correct probleme of comm world on SuperLuDist (complex version)
    - correct medit Makefile.am in case of no compilation of medit ..
    - correct a lot of MPI parallel solver example 
        a bug pastix interface ???
	MUMPS, superludist, hypre , hips works 
    - correct link problem in hips and hypre under linux
    - Add thresholdings.cpp	thresholdings.edp in examples++-load to remove to small coef
       in a matrix . 
    - Add lots of DDM Schwarz GMRES precondionned with a coarse grid solver por overlapping : 
      DDM-Schwarz-Lame-2d.edp		DDM-Schwarz-Lap-2dd.edp		DDM-Schwarz-Stokes-2d.edp	
      DDM-funcs-v2.idp DDM-Schwarz-Lame-3d.edp  			DDM-Schwarz-Lap-3d.edp			
      DDM-Schwarz-macro.idp
      I will add the explanation in the doc in the future; 
      

    - Add a true exemple to buld mesh form a image (lg.pgm) ,  Leman-mesh.edp 
    - Add New syntaxe in macro generation
        NewMacro a(i) 
        EndMacro 
       with // comment and macro definition, first trik
     - Add interface with special function  of gls  http://www.gnu.org/software/gsl/
        the full list is in the example examples++-loal/gsl.edp  
     
v 3.11-1     25 dec 2010
     - Add coarse preconditioner in MPIGMRES[23]d.edp (to be optimal)
       now the number of iteration is close to 10.  
     - Show ff++ line number in case of  assertion  in RNM class. 
     - add Coarse Preconditionner for  MPIGMRES[23]d.edp (Good)
     - solve bug For MPIGMRES2D.edp  due to interpolation bug in rare case
       by Add brute force for seach of point in 2d like in 3d in the search in find ouside 
       set global variable :  searchMethod=1;  // more safe search algo (can be expensive in case of lot of outside point) 
      -   add hack of ILU precond, if tgv is < 0 then
        we remove all the line and put 1 on the diag term ..
v 3.11 9 dec 2010 
   - update the documentation of 3d adaption process
     and / correct  freeyams,mshmet, mmg3d  interface and associaded  example
   - configure is compatible with MKL lib (on gnome)
   - add quoting argument in macro argument with { } for mpi  plot .. 
v 3.10-2
   - try to compile with MKL libs.
   - correct MPIGMRES[23]d.edp example 
     add doc on this example.  
v 3.10-2
 - add operator to inverse permutation to set or initial int[int] array
   I=J^-1; 
   if J is a permuation of 0:n-1  then we have :   I[j[i]]=i 
 - correct comment problem of periodic boundary condition in 3D
    see examples++-3d/periodic-3d.edp
 - correct configure to scotch compilation  (phtread) 
v 3.10-1
  - at convection function form formal array to int,real,complex array
   (resp.     toZarray, toRarray, toCarray)
  - correct ffrandom.cpp to read /dev/random to get a true random seed
    srandomdev 
  - add ff-mpirun script to simplify the launch of FreeFem++-mpi version
  - correct MPI for MPI icc on gnome 
  - correct pb of computation of area, lenbord in mesh type in some case .
  - correct WHERE_LIBRARY-config  for blas find in configure
  - correct atof 
v 3.10 
  - add true Domain Decomposition example in 2d and 3d. 
      see    MPIGMRES2D.edp	  MPIGMRES3D.edp in MPI  examples 
  - add mpi Isend/Irevd for complex struct like matrix, meshes 
     for send/recv Huge objet 
  - add named parameter verbosity= in add GMRES and GC function 
     for show algo evolution .
  - add MPIGC dynamic LIB for // GC and GMRES , add fully // scharwz RAS
     in MPIGMRES2d.edp 
  - correct  problem of  a=A^-1*b when a, or b is not consecutive array
   ie. expression like  M(2,:)  
  - correct problem of Makefile in download  clean,install,WHERE target
v 3.10 ( 5 oct 2010)
  - add install of missing MPI  dynamics lib. 
  add argument  "-cd"  to FreeFem++ command to change  current directory  to edp script directory 
  - clean configure.ac 
  - do correction for g++-4.6 compiler (lambda expressions , and trap ..)
  - add  MPICG  for Parallel Conjugate Gradient for full split matrix
     see MPICG
  - correct Makefiles in download ( add WHERE interface) 
  - remove wait option in medit because the code is wrong , always waiting  now 
  - correct bug renumbering  of matrix :  B= A(I,J),
    the last term N,M was force to zero  =+ -> += (line 1885 of file lgmat.cpp) 
    where N= I.max, M=J.max 
v 3.9-3
  - correct compilation of gmm, mumps on linux
  - add parameter -ne in FreeFem++ commands  to remove edp script print
  - correct mistake in --enable-m64 or --enable-m32 (suppress the configure warning)
  - correct of ff-c++ script if whith space in path (for windows)
  - add compilation  of gmm library 
v 3.9-2
  - correct compilation of mshmet dynamic library
  - correct pb of compile of superludist ( add CNOFLAGS no optimize CFLAGS)
  - correction of lib on win32 for freeyams and mmg3d
  - correction download/Makefile.am to be sure than bin is a directory.
  - add fftw of win32 

v 3.9-1
  - correction  mmg3d interface (J Morice) 
  - correct of mmg3d and freeyams under wind32 (ld problem)
v 3.9-1  August 2009   (For FreeFem++ days )
  - correct configure (find lapack lib change $ll_lapack_libs in $ll_lapack_lib)
  - correct mistake in mpi 
    add:  gatherv, allgatherv, scatterv, alltoallv   
        and complex data type in:  allgather, gather, scatter, alltoall, ..
    correction bug : gather,  scatter
    correct essai.edp mpi example 
  - correct problem of compatibility of dynamic lib  with and without MPI 
     change ff-cc++ to add all MPI libs if MPI version exist. 
  - correct default  region number in square build mesh function  now 0 , (2 between  v3.8 -- 3.9)  
  - change in all example  reffacexx= in labelxx= ..., etc to be correct with v 3.8 change.
  - correct a mistake of type  region= parameter in tetgen all functions.  
  - correct the mpi configure search tool
      miss when no full path are given  ( only change in configure.ac) 
v 3.9  July 2009.
  - add lots of automatic  compilation of download soffware,
   
	tetgen superlu fftw metis yams mshmet 
	blacs parmetis scalapack scotch superludist MUMPS pastix hypre hips 

        For the link with mmg3d software put the tar.gz archive in .../dowload/pgk directory.
    add interface with freeyams, mmg3d, mshmet (3d mesh adatation) software,
    add automatique compilation of // solver with 
      flags -auto in ff-c++ commands
        

v 3.8-2
  - add cast operator from SubString to String 
  - correct the SubString tools 
v 3.8
  - add in change mesh2 tool to make a renumbering of vertex for periodic 3d mesh. 
     Th= change(Th,renumv=old2new);
  - correct ' operator do alway in complex case the conj and trans (Hermitian stuff)
     in  formal array with [ ]' 
     in  linear comb of complex matrix  
        now  A + (-1)*A' is zero  is A is hermitian .
     in varf term  (a*dx(u)*dy(v) )'  <=>  a' dx(u)*dy(v)
    now you can  [a*dx(v),dy(v)] ' *[a*dx(u),dy(u)] in complex case.. 
  - plot of complex field  and 3d vector .. 
  - upgrade the documentation in 3d mesh example
  - change configure.ac for mpifc, mpif77,  mpicc, ...  
  - correct  include seach path 
  - uniformize named parameter  in change, movemesh, in load "msh" , glumesh, ...
     add synonyme:  
        refface=  ->   label=
	reftet=   ->  region= 
     in buildlayer:
	labebup=
	labebdown=
	labelmib=
  - writing  schwarz-nm-3d.edp examples
     - add array of 3d mesh
     - add  word  volume to get the volume of the current tet element
     - correct  metis.cpp in 3d case 
  - correct in fflapack.cpp (example++-load, computation of eigenvalue of complex matrix)
         -llapack search in configure.ac 
        and unify fflapack.cpp lapack.cpp are the same file.
  - add seekp, tellp  method  on ostream type 
        seekg, teeg   method on istream type
        see examples++-tutorial/readmesh.edp examples 
v 3.7-1 13 jan 2010.. 
  - EqPoisson.edp (Solve Poison equation in 3d fish)
  - add possibility to put array  (int[int] )  to set a set of label/region in
    in integral (int1d,int2d,int3d)   or  "on" key word 
    example : int[int] l=[1,2,3]  ,  ...  on(l,u=1);
  - add tool to compile under mingw/msys under windows 32 (freefem++-nw works, 
  -  add bug (random trap) in bamg in sub domain computation in case of 
    internal edge in    same region (fist time in 15 years). 
  -  add bug in varf interpretation in  complex case (miss some convertion)
  - add build  interpolation matrix in 3d (see examples++-3d/mat_interpole.edp)
  - correct CFLAGS (add -fFIP) in superlu , umfpack under 64 architecture. 
v  3.7   8 december 2009 
   add 3d beam examples examples++-3d/beam.edp 
   correct install problem due to metis directory
   add dynamic load interface with newuoa fortran optimizer without derivative
      see ffnewuoa.edp example ins examples++-load
   correct problem of free of mesh in case of  gluing meshes
     Th= Tha+Thb; // now just the ref counter on Th  is decrease
                  // and  before  Th is delete 
   add .im, .real method on complex [int,int] matrix array 
   add  missing  matrix<complex>[int]  type in grammar.

   correct mistake in generation of file WHERE_LIBRARY-config
   add  initialization of	scalar variable at zero 
   add  warning in case of use of	variable with hide freefem++ name
   add missing file in download version

v 3.6.1 
    correct  font problem in documentation
             lot of mistake in examples++-load 
    add tool to compile all examples 
    add PICHON stuff for brute force search point
    add umpack long interface to by pass the 32bit limits 
      see UMFPACK64.cpp and LapUmfpack64.edp in examples++-load.
    add some interface with lapack (inverse of full matrix, 
       eigenvalue of full matrix) 
    correct  ::Draw undef ref.  ubuntu 10. (gcc 4.4)
    correct typo error  image tools
    see UMPACK64  (code and example in examples++-load)
v 3.6 (4 nov 2009)
    change the version of tetgen to 1.4.3 download
    correct in ffglut max value of arrow
    add examples of function with stack parameter in 
       examples++-load/funcTemplate.cpp load facility
       and 
       examples++-load/funcTemplate.edp
    add tools to read .pcm files for optic-flow computation (pcm2rnm.cpp) 
  	 see http://www.cs.otago.ac.nz/research/vision/Research/OpticalFlow/pcm.html
     correct bug in GMRES without preconditionner assert error. 
     correct a rare and ramdomly bug in the interpolation 2d process when the
      the element 0 touch the boundary
     correct on versy old bug in bamg in case of internal boundary edge with
     two extremity  in true boundary (symptoms FillHoleInMesh trap).
   correct operator  subsript a:b:c  operator (in some case the last value was miss) 
   correct the  ?: operator in case of constant array  
  - add -enable-m32 flag to build 32bits freefem++ on 64 architecture computer
  - correct eigenvalue.cpp in case of complex finite element
v 3.5-2  (28 sept. 2009)
  - correct install problem under windows (metis.dll do not compile)
  - correct MacOS install missing /usr/local/lib/ff++/3.xx-yy link 
v 3.5-1  (24 sept. 2009) not to bad ...

   - add interface to metis 
   - correct  display of 3d mesh (with light today)
   - add trap in case of mesh with negative volume 
   - correct bug  in case  3d finite element with constant number of df / node
     example fespace Vh(Th2,[P1,P1]); or fespace  Wh(Th3,[P0,P0,P0 ])

   - correct problem with tetgen under windows and ubuntu 
       (remove optimization during compilation of the library)

   - correct setting array of vectorial finite element Wh 
     Wh[int] [u,v](10); // array of 10 value
     u[1] or v[1] 2 componante function for indice  1.
     [u[2],v[2]] = [1,2]; //  set the function for indices  2. 
   - correct in matrix size of  matrix B=A(I,J) to I.n x I.m 
     where I, J are array of int 
     
   - add resize of complex sparse matrix 
   - Add inferface with metis in examples++-load   (metis.cpp, metis.edp)
v 3.5 (27 aug. 2009)
  Complet writing to MPI interface, tested with Open MPI 1.2.3 and mpich 2.1 
      see chapter 10, page 229, of  the freefem++doc.pdf  and 
      see example++-mpi/essai.edp.    

  correct 
      operator  a(:)= 1:4; // before change a copy of a and so do nothing
      string[string] s; 
      s[i]; // before trap (two delete) 
v 3.4-2
  add boundary mesh inquire in 2d and 3d
  Th.nbe ; // return the number of boundary element 
  Th.be(k);   // return the boundary element k $\in \{0,...,Th.nbe-1\}$ 
  Th.be(k)[l];   // return the vertices l $\in \{0,1\}$ of  boundary elmt k 
  Th.be(k).Element ;   // return the triangle containing the  boundary elmt k 
  Th.be(k).whoinElement ;   // return the edge/face number of element containing the  boundary elmt k 
  Th[k].adj(e) ; // return adjacent element to k by edge/face e, and change 
  // the value of e to the corresponding edge in the adjacent element
  Th[k] == Th[k].adj(e) // non adjacent element  
  Th[k] != Th[k].adj(e) // true adjacent element 

  do small change in msh3.cpp  file 
 
v 3.4-1
   rewrite of the mpi interface in vue of use group and communicator
     now we only use mpi.h and not mpi++.h
   add plot of array of mesh and array of finite element functions
   add new command in ffglut graphic interface
     n,N,i,I  to change the number of iso-value and to set the min,max value to the
           graphic
   add left button motion to zoom , and + alt to translate
V 3.4
    add in parallel version mpi send, receive and broadcast of  3d mesh and matrix ( sparse matrix).
    build a 3d mpi schwarz example (see example++-mpi/schwarz-3.edp)
    make a wrapper (launchff++.exe)  to launch freefem++ under windows/OS in the edp file directory. 
    correct in interpolation matrix ( unset variable in case of same mesh 
      => missing some term in the matrix) 

V 3.3-3 
  add missing file in distribution examples++-load/DxWriter.cpp
V 3.3-2  8 June 2009
  add   2 new finite element:
   Edge03d : the Nedelec 3d  Edge finite elements.
   RT03d : the Raviart-Thomas 3d  finite elements
   correct the size of anytype for 64 bits architecture.
   install the new .idp file (include of edp file) in
   directory idp    

V 3.3-1 1 June 2009
  to correct the problem of the destruction of parameter in case: 
      func real[int]   a2(real[int] a) {a[0]=2; return a;}
   so now the parameter cast to KN_<double> so  the array is not
   duplicate like in C++, and the value of the parameter is change. 
   
  - simplify the internal (C++)  definition of array now of 2 type
    KN<> * of variable (left expression)
    and  KN_<> of   right expression 
  - correct UnRef<A,B> where B ~ A*

v 3.3  29 may 2009
  -  correct some problem in return in function :
   the following function prototype now work without memory fault. 

    func real[int]   a1(int n)  {real[int] a(n);return a;}
    func real[int]   a2(real[int] &a) { return a;}
    real[int] G=[7,8,9];		  
    func real[int]   a3(real[int] &a) { return G;}
    func mesh carre(int n) { mesh th=square(n,n);  return th;}
   (lot of change, warning:  now KN<K> is not a freefem++ internal type). 
  warning : 
  -  in int3d flag qfV=  a 3d quadrature formula
        the 3d quadrature formula are
         qfV1 , qfV1lump, qfV2 , qfV5  (where the number is the order of the quadrature)
  -  correct sign of the normal in 3d (now the normal is exterior)
      -------------------------------------------------------------
  -  correct in adapted mesh the flag splitin2=  (no tested)
  -  correct the convect operator in 3d 
  -  add EigenValue tool  for 3d case  May 15, 2009. 

v 3.2
  - add tools to set array like in matlab/scilab
    real[int] a(1:2:10),  def  a to 1,3,.. 10.
    int[int] I(0:n-1) , set array of size n to 0, .., n-1
    real[int] b(0.1:0.5:9.99) , set   0.1, 0.6, 1.1,..., 9.6  
    add :   b.n == (9.99-0.1)/0.5 +1
  - sort(a,p);  // sort a is and array and p in parallel 
     where is a integer array same size of a.
        
  - add interface with MUMPS  parallel solver (in progress) 
    http://graal.ens-lyon.fr/MUMPS/

  - rename  movemesh2D3Dsurf in movemesh23 and change named 
     parameter normal= in orientation=

  - add periodic boundary in condition like in 2d
    example for a cube with label face numbering 
      1 :  ( x == xmin)        2 :  ( x == xmax) 
      3 :  ( y == ymin)        4 :  ( y == ymax) 
      5 :  ( z == zmin)        6 :  ( z == zmax) 
     fespace Vh(Th,P2,periodic=[[1,y,z],[2,y,z],[3,x,z],[4,x,z],[5,x,y],[6,x,y]]);

  -  automatic: Finite element name conversion  beetween 2d and 3d
      old : fespace(Th3,P13d);
      new : fespace(Th3,P1); 
  - add complete rewrite of the way  to set parameter of the sparse solver
    for  parallel solver ( no progress)
  - add --enable-m64 for 64 bit compilation flags 
  - correct f2c of  take the -m64 flags 
  -  build the universal f2c library with 4 architecture ppc i386 ppc64 x86_64

v 3.1-1
  - add resize of sparse matrix ( only  morse matrix)
  - add  new 3d finite element P1b3d   (P1 bulle 3d) 
  - add  new option in int1d function in case for  build matrix with different meshes
        mortar=1 ( do not remove integration point if the point is outside of the domain)
    to make no to bad integration on same curve with different meshes. 

v 3.1
  -correct compile  order missing libMesh
  - add lighting in ffglut (key l) for 3d isovalue 
v 3.0-6
  - correct int2d in 3d. bug in   Face -> tet mapping  (PBord method) March 5 2009
        change the postscript  plot dimension to be in A4 paper.  
        nuTriangle   given the tet  number in 3d case (5 march 2009)
        area  given the area of the triangle on 3d border integral. 
  - add tool to required edge in adaptmesh with label of edges. 
  - correct  dy(uh) (always 0) for P13d finite element  (27 jan 2009)
      op==op_dy =>  op==op_dy line 97 P012_3D.cpp
v 3.0-5 (20 jan 2009)
  - correct array 3D  FE function +  missing function in 3D (not finish) 
  - correct the install on under windows / cygwin 
      add src/Graphics/ff-win32.cpp missing file in tar.gz under windows
      correct ff-c++  under cygwin compile (fist test) 
  - automatic compilation of BLAS, suppress the automatic compilation of BLAS atlas (to long) 
  - ffglut improvement 
        - now save 10 graphic states and use "p" key to show previous graphic
	- do automatic z rescale on Z direction the  3d plot 
v 3.0-4 (05 jan 2009)
  - correct int2d in 3d case: all this integral was multiply by 2. 
  - add 3d viewing in plot (in test)
v 3.0-3 (27 dec 2008)
  - build a not to bad version 
  - remove so crazy print 
  - change the ffglut  io with  freefem++  ( to be more universal independent of the architecture an more universal)
    - correct ffglut in case of plot of array and when the plot is empty.
      remark ffglut is not compatible with previous version (sorry)
    the filename create with option w is ffglut_xxxx.ppm (window dump)
    correct int2d on 3d mesh3d for Neumann Boundary condition
v 3.0-2 ( 9 dec 2008)
  - ff-c++ to dynamic load program (add ff++.hpp file  with all include, may be)
  - correct the Makefiles et load-link, 
  - correct ffmedit (binary read /win32)
  - cosmetic change in ffglut + stabilization
v 3.0-1 (4 dec 2008)
  - complete change of the graphic (freefem++ is now the old freefem++-nw ) 
      - add  medit (a visualisator software by  P. Frey)  inside freefem++
          with a name ffmedit  
      - suppress the ide version
          see page: http://www.ann.jussieu.fr/~lehyaric/ffcs
      - We put a client-server architecture
        (freefem++,freefem++-nw)  <->  ffglut
	- freefem++-nw and freefem++ is now the same file 
	- ffglut is the visualisator in glut library (run on linux, mac, win32)
    how to use now: 
        freefem++  file.dp # by default load ffglut (so ffglut must be in the PATH).
	freefem++  -glut ffglut-path  file.edp  # change ffglut command
	freefem++  -fglut saveglutfile  file.edp # save data plot
	ffglut saveglutfile # plot a data file
	freefem++ -nw file.edp # freefem++ with  graphics 
	ffmedit  #  used medit in freefem  
	

  - now glumesh2D dynamics tool is in freefem++   
        add operator + beetween meshes with glu
        examples++-tutorial/glumesh.edp
        
  - add 3 meshing tools in msh3.{so,dll,dylib}      	
	operator + between mesh3
	add freefem++ function:
	  movemesh2D3Dsurf movemesh3D buildlayers
        see:
    	  examples++-load/buildlayermesh.edp
	  examples++-load/glumesh3D.edp
	  examples++-load/refinesphere.edp
	  examples++-load/tetgenholeregion.edp

	    
  - add interface with tetgen  3d mesh generator (tetgen.{so,dll,dylib})
        add freefem++ function: 
	  tetgconvexhull tetgtransfo tetg tetgreconstruction
	see:
      	  examples++-load/buildlayermesh.edp
	  examples++-load/refinesphere.edp
	  examples++-load/tetgencube.edp
	  examples++-load/tetgenholeregion.edp

  - add interface with medit (medit is now in freefem++ this call
       ffmedit in medit.{so,dll,dylib}
       add freefem++ function medit savesol (2d , 3d  medit plot)

  - correct ?: operator with array and real. 
  - correct memory error with  valgrind tool (19 oct 2008)
       In ~ConstructDataFElement() change counter after free
       nosym eigen call to neupp missing +1 in size of array 
  - correct gestion of the counter of ConstructDataFElement class (09/2008)
       in Valladolid spain. 
  - add 3D FINITE ELEMENT P0,P1,P2 
     mesh3 Th("toto.mesh")
     int3d(Th) , int2d(Th)
     see examples++-3d/first.edp example
    
v 2.25
  - add tool to change lab in mesh see  glumesh.edp examples++-load 
v 2.24-4
  - add missing code in case of DG (jump or average) of test function in
    linear form
  - add tools to glu meshes (not well test) see glumesh.edp in load examples.
  - change definition pour interpolation point for P1dc and P2dc  (P-G)*c+G
    now all interpolation point are fully inside the triangle 
v 2.24-3
  - add option resid= in eigenvalue function (see arpack doc)
  - add  formal operator  ./  .*  between two  [ a , b , c ] array
v 2.24-2
  - correct convect explain in documentation. 
  - correct configure.ac build FLIBS and add check of FLIBS, X11_LIBS value
  - correct buildmesh trap when degenerate border is create  by error.
v 2.24-1
  - correct periodic condition  in case of one edge 
  - correct doc file
v 2.24
  - bug with border and func
     this kind of code now works
       func abc=  a(6) + b(4) + c(4)  ;
       func def = d(4) + e(4) + f(6);
       func bbb= abc  + def;
      plot(bbb); mesh Th=buildmesh(bbb);
     whare : a,c,b,c,d,e,d are  border. 

  - correct bug in periodic Boundary Condition( find by Alexandre Masserey <masserey at ycoorsystems.com>)
       mistake in abscise choose.
       sometime assert fail (loop).
     add option to the buildmesh that the boundary is fixe (fixeborder=true)
     see : examples++-tutorial/periodic4bis.edp   
  -  and  R1.hpp, R2.hpp, R3.hpp files for the 3d version.  
v 2.23-2
  - correct trap in adapdmesh (in mortar-DN-4.edp) in very rare case 
    du to  initialized   adj  of edges on geometry (very old bug) 
    in case of multi points.
  with verbosity=10; you get this message: 
 -- Begin of insertion process 
 bug  2
 Bug double points in 
....
 Fatal error in the meshgenerator 5
....

v 2.23-1
  - correct assert fail in check in case of no UMFPACK
v 2.23
  - add new edge finite element (P1 to P5) 
   in examples++-load/Element_PkEdge.cpp examples++-load/testFE-PkEdge.edp  
  - preconditionned  mortar example Neuman -> Dirichet in scalar and parallele
    see examples++-mpi/mortar-DN-4-mpi.edp examples++-tutorial/mortar-DN-4.edp
  - add vectorial  operator  d=  a ? b : c  where d,a is a array, and b,c can be array or scalar 
    usefull for boundary condition
  - correct problem in mpi recieve message
  - correct load dynamic on MacOs (bug in fortran find library) in configure.ac
v 2.22-2
  - small change in configure.ac of find libumfpack on debian system
  - correct checking code in problem to same the mesh must be unique in problem of solve definition.
v 2.22-1
  - make change in Makefile to build universal binary on MacOS
  - change the README file for compilation on ubuntu Linux.
v 2.22   (december, 18, 2007)
  -  correct problem of string argument in function
  -  correct all elasticity problem (missing sqrt(2) )
  -  correct save of log file under  Windows XP, ..
v 2.21   (november, 18, 2007)
  -  correct documentation for new edition
  -  correct configure for no X11 and UMFPACK in sparsesuite
v 2.20-1 (october, 24, 2007)
  -  correct compilation problem without umfpack
  -  add  formal operator  for vectorial operator in variational form
        +, -  on  [  ]  array
        trace operator  matrix like [ [ ... ], .., [... ] ] 
        take element of an array [ ...][2] 
v 2.20  (october, 2, 2007)
  -  add in download automatic generation of f2c and fort77 to get a free  fortran compiler
	just  go in 
	(download/f2d;make install)
  -  add automatic find of libf2c in case of F77=fort77 (driver of f2c) in configure 
  -  add tools to change the sparse linear solver with load dynamic, see doc and SuperLu example
     in examples++-load
v 2.19   (august, 19  2007)
  - correct buildmesh from geometry file and add nbvx= named parameter
    to infore the maximal number of vertex in a mesh.
  - add  Th[k].area to get the area of the k-th triangle of Th.
  - add  Th[k].region to get the region number (label)  of the k-th triangle of Th.
  - correct string delete in plot of border (mesh.edp)
  - correct mean and jump in  interpolation operator.
v 2.18-1 (august, 2007)
  - correct old bug under Windows: the std input stream  now works (cin) 
v 2.18  (july, 19, 2007)
  - add sort of real array or int array, syntax: 
      a.sort, and we have : a[i-1] <= a[i] for i =0 to a.n-1
  - add  v=a.quantile(p) ;  statistical function  and commute v such than 
     #{ i / a[i] < v } = r*a.n   <=>    v = a[r*n]  when the array is sorted.   
  - correct P4nc, P4 finite element
v 2.17-2 (19 june 2007)
  - correct double   $(BLASLIB) in configure.ac (one more)
  - correct old bug in bamg, in case of bogus boundary (auto
    crossing)and   add code  to  handle  this error correctly with exception
    (see "test to catch bogus boundary"  of example++-tutorial/mesh.edp).
v 2.17-1
  - add the command line parameter [-v nn] in all freefem++ program to set 
    the level of verbosity to nn  before all, and if nn == 0 then theoretically no more
    spurious output.
  - correct  automatique choose of  color in plot of arrow. 
v 2.17
  - add bessel function   j0, j1, jn, y0, y1, yn -- Bessel
        functions of first and second kind
  - add  erf, erfc -- error function operators
  - add   tgamma, lgamma  -- gamma and log of gamma
  - add   tool set and get   line,column and value of a sparse matrix
             [I,J,C]=A; // resizing  array I,J,C
             A=[I,J,C] ;
            where
             int[int] I(k),J(k); real[int] C(k);
  - add   in tools to get eigen vector of None FE problem
   like:
     int nev=5;  // number of computed eigen value close to sigma
     real[int] ev(nev); // to store nev eigein value
     real[int,int] eV(AA.n,nev);   // to store nev eigen vector
     int k=EigenValue(AA,BB,sym=true,11value=ev,rawvector=eV);

  - correct the quadrature formula automatic choose with the qforder= field
  by default the order is correct until  order 10 (so exact for P9 polynomials)
  to add new Quadrature formalur up two  order 26 add  dynamic load qf11to25
v 2.16-2
  correct configure.ac for windows ide compile
  correct meshsplit, trunc  function 
v 2.16 april 17 2007
  Change umfpack download version to the last one
  add string input from file or cin
  add Compressible Neo-Hookean Materials exemple
v 2.15-1
  Correct mistake in cas of splitmesh with internal boundary (assert)
  Add dump  of the mesh regulary (personal work, FH)
v 2.15  (27 mars 2007)
  The blas.tgz file change on http://www.netlib.org/blas/blas.tgz
  correct the makefile  (remove BLAS dir) 
v 2.14-3 (24 mars 2007)
  Correct bug in vectorial operation of type  (b - 3.14*d) (Thanks to F. Dortu)
v 2.14-2 (22 mars 2007)
  correct integration problem on none classical mesh (mesh of curve of mortar technique)
  the mesh::Find methode is boggus of the kind on meshes, i make an optimization to remove
  this call in case build matrix with varf int1d(Th)( u*v) , in case of Th, u or v is def.
  on same meshes.	
v 2.14-1 (0 mars 2007)
   correct missing speeling in  configure.ac  mpi++.h -> mpi++.h for mpich.1.2
   add computing linear form of 2 meshes  (ex. v on mesh Th1, and  compute int2d(Th1)(v) )
v 2.14  (7 mars 2007)
   make correction for solaris bluid
   add P3,P4 and new quadrature formula on triangle up degree 25 
   with dynamics load. See examples++-load/*P[43]*.edp
   correct mistake when we build matrix from varf with 3  meshes (FH).
v 2.13
   Add formal operator * and ' (to day just transpose) 
   on vector [ .. ]  and matrix [ [],...,  []]
   so we can write  
    macro grad(u) [ dx(u),dy(u)] // EOM
    int2d(Th)( grad(u)'*grad(v) )   
    add scalar term in block matrix as 1x1 matrix.
v 2.12-2
  correct problem in splitmesh and trunc
  trunc build always a part of a subdivision mesh
  splitmesh given more pretty mesh, before if we split 2 adjacent
  triangle resp in 3 and 4, before we put 2+3 points on the common edge
  now the put 3 points. 
v 2.12-1 (jan 15,2007)
  correct early delete of return pointer in freefem++ function (trap in string.edp under windows )
v 2.12   ( jan 10,2007)
  correct mistake in none square block matrix 
v 2.11-2 (nov 29,2006)
  correct integration problem on none classical mesh (mesh of curve of mortar technique)
  the mesh::Find method is bogus of the kind on meshes, i make an optimization to remove 
  this call in case build matrix with varf int..(Th)( u*v) , in case of Th, u or v is def.
  on same meshes.  
v 2.11-1 (nov 28,2006)
  Correct bug in interpolation, we building fespace Wf(Th,[RT0,P0]); 
v 2.11 (nov 10,2006)
 Correction of  the problem in arpack (eigenvalue computation) under i383 processor.
 After 2 week of hard work, a found the bug.
 In new version of LAPACK second is a real function
 and in ARPACK second is a subroutine  (Thank the fortran).  
 So I change the Makefile to change the name of function second into secnd2. 
 How to find the bug, in this case the stack of the floating register (info float under gdb) 
 are growing. So after any real function can return a NaN (one time) due to stack float register 
 overflow.  A big Thank to D. Bernardi of the help (Merci au fortran et Merci au i383).

v 2.10 (oct 26,  2006)
  correct problem with Eigen value on window and some linux distrubition (Fedora)
  add a patch of two functions in lapack  see  arpack/arpack-patch-lapack.tar.gz

v 2.9 (sept 19, 2006 FH) 
  change Send and Recv in mpi  to Isend and Irecv to make async communication
  so now freefem++-mpi work with openmpi.

V 2.8-3
   correct freefem++-cs trap under windows (install the static version)
   add tools to format ostream  ( file or cout)  

V 2.8-2 
   (july)
   use f2c as a fortran compiler on  MacIntel architecture,
    correct mistake in mpi. 
   (june 2006)
   add nint, and correct mistake in ceil function
   add configure flag to remove cadna --without-cadna
V2.8
   Correct mistake in OtherMacOsLib.tgz store file and not link (after 1 june)
 * Correct bug in vectorial problem with same approximation, some time
   we do a renumbering of the previous  approximation space (a pointer mistake).
V2.7-1
  * change size of data in 64 bits architecture (in file AnyType.hpp)
V2.7
  * correct mpi Makefile
  * correct bug in construction of recursive composite finite element (see HISTORY)
    remark: the finite element struct, so all dynamic library must be recompile. 
V2.6-1
 * add the examples-bamg  directory  in freefem++ and  correct bamg graphic
V2.6
 * Huge work (more than a full week) to improve the windows version 
   change all the .dll construction
   remove lots of trap. All example run without graphics
V2.5.1
 *  change nbve= to nev= in EigenValue function  (to be compatible with the doc)
 *  newtow to newton    (to correct misspelling)
 *  change in AnyType to remove da random bug on windows.
 *  vim color syntax from Richard MICHEL (vim_highlighting_for_FF++.tar.gz)

V2.5-0 
  *  add C++ string operator :  int i; string s = "....";
       i=s.length ; i=s.size; 
       i=s.find("qsdqs");i=s.rfind("sqs"); i=s.find("qsq",2); 
       s(3:4)="sdfffsf";  // replace from char 3 to 4 by 
       // s[3] not alloued to day (char type is not well type in freefem++).
       getline(cin,s); 
  *  Correct bug when passing parameter string in function
  *  add freefem++  mode for mi editor on macos http://www.mimikaki.net/en/
    store in  mode-mi-edp.zip archive (unzip and put in the Folder  opened with the 
    mi "Option->Open Mode Folder" menu and set "mi" is the Default application 
    for all the .edp file). 
    
V2.4-2
  * add lot missing vectorial operation with sub array 
  * if u is a complex array (i.e. complex[int u] ..;) 
     then  u.re (resp u.im) are  the real array of the real (resp. imag ) 
     part of the vector u
  * in block matrix add utilization of array to build the block matrix
    see new Laplace-lagrange-mult.edp file to have an example
V2.3-3
 * add outer  product to set matrix or full real or complex 2D array
    A = 2*a*b'; // where A is a "real[int,int]", b and c a "real[int]".
    A += 2*a*b'; or  A = 2*a*b';
   add matrix and array renumbering (see sec. 4.8 of freefem++.doc for all details)
    
V2.3-0
 * add three key word for future use try,catch throw 
   a exception handling (no memory management.) 
   try {   code ; }
   catch (...) { code ; }
 * use tolpivot= to set the pivottol in LU, crout, cholesky factorization 
V2.2-1
 * Add init file to set verbosity, includepath  , loadpath and 
   add preload (dynamic link file)
   the file are under unix and MacOs
	 /etc/freefem++.pref
         /Users/hecht/.freefem++.pref
         freefem++.pref
   under windows
         freefem++.pref
   and the syntaxe of the files is

     verbosity= 5
     loadpath += "/Library/FreeFem++/lib"
     loadpath += "/Users/hecht/Library/FreeFem++/lib"
     includepath  += "/Library/FreeFem++/edp"
     includepath  += "/Users/hecht/Library/FreeFem++/edp"
     #  comment 
     load += "funcTemplate" 
     load += "myfunction" 

 * Add search directory for include, and load  under Unix and Windows
   Just on the shell sh  set environment variable
    export FF_VERBOSITY=50 
    export FF_INCLUDEPATH="dir;;dir2"
    export FF_LOADPATH="dir;;dir3""
   remark the separator of directory  is ";" and not ":" because ":" is use under Windows.
 * Examples of Add new Finite element with dynamic link/load,  Morley and Bernardi Raugel   
   see examples++-load 
V2.1-2
 * add more example of dynamic load example (thank to F. Dortu)
 * correct mistake in simple integral computation (not in variational 
   form)  before we alway use a quadrature formula of order 3, 
   now we can change the quadrature formula and the default is of order 5. 
V2.0-4
 * change in mshptg the size of the small possible edge from
   1/32000 to 1/1 000 000 000, this mesher is use in trunc, 
  splitmesh, and triangulate functions.

V2.0-3
 * correct operator x =/,  it does  nothing before. (v2.0-3)
   see end  examples++-tutorial/array.edp for all array operator
   =  +  -  *  /  .*  ./ += -= /= *=  : .l1 .l2 .linfty .sum .max .min ' 
   
 * correct bug when solving complex linear system with UMFPACK
   before we solve the system when the conjugate matrix


Major Change previous version
-----------------------------
 * add ternary C expression
     Symbol       Example      Explanation
       ?:          a?b:c     ternary operation
 
 * add 
   sparse matrix computation,
   block matrix construction 
   interpolation construction
   see 
     examples++-tutorial/sparse-cmatrix.edp  
     examples++-tutorial/sparse-matrix.edp
     examples++-tutorial/mat_interpol.edp
   not so well tested, but could be worse. 
   
 * add example to solve variational inequation. VI.edp

 * add  possibility to build matrix where the 3 meshes 
   the integrale mesh, unknown FE mesh, test FE function 
   can be different.  You must use varf to build the matrix 
   is not a standard problem.
  
 * Improvement of  dynamic loading facility under the 3 systems
   add a fast Fourier transform with dynamic loading
   see examples++-load/dfft.edp

