[Gmsh] compilation problem with gmsh version 1.60.1

[Christophe Geuzaine]

Romain Quey wrote:
> Hello,
> 
> I'm still using gmsh version 1.60.1 for some applications. With the
> sources available from file 'gmsh-1.60.1-source.tgz', I've got
> some problems to compile the program (no problem for any later version
> of gmsh). I'm on a standard linux pc (fc4).
> 
> './configure' seems ok (see attached files), but 'make' generates a
> first error for '2D_Mesh_Triangle.cpp' (see make.log), and the same one 
> for 'Triangulate.cpp':
> 
> Here is the error:
> g++ -g -O2 -DHAVE_GSL -DHAVE_MATH_EVAL -DHAVE_NETGEN -DHAVE_TRIANGLE 
> -DHAVE_LIBZ -DHAVE_LIBPNG -DHAVE_LIBJPEG -DHAVE_FLTK 
> -I/usr/local/include -I/usr/X11R6/include -I../Numeric -I../NR 
> -I../Common -I../DataStr -I../Geo -I../Mesh -I../Graphics -I../Parser 
> -I../Fltk -I../Triangle -I../Netgen -I../Netgen/libsrc/include 
> -I../Netgen/libsrc/interface -c 2D_Mesh_Triangle.cpp
> ../Triangle/triangle.h:285: error: variable or field 'trifree' declared 
> void
> ../Triangle/triangle.h:285: error: 'VOID' was not declared in this scope
> ../Triangle/triangle.h:285: error: 'memptr' was not declared in this scope

That's a problem with new versions of Triangle: to fix it, just add
"#define VOID void" after "#define REAL double" in
Mesh/2D_Mesh_Triangle.cpp.


> 
> The other error is:
> g++ -g -O2 -DHAVE_GSL -DHAVE_MATH_EVAL -DHAVE_NETGEN -DHAVE_TRIANGLE
> -DHAVE_LIBZ -DHAVE_LIBPNG -DHAVE_LIBJPEG -DHAVE_FLTK
> -I/usr/local/include -I/usr/X11R6/include -I../Common -Ilibsrc/include
> -Ilibsrc/interface -DNO_PARALLEL_THREADS -UWIN32 -c
> libsrc/meshing/global.cpp -o libsrc/meshing/global.o
> libsrc/meshing/meshtype.hpp:965: error: ISO C++ forbids declaration of 
> 'Mesh' with no type
> libsrc/meshing/meshtype.hpp:965: error: expected ';' before '&' token
> libsrc/meshing/meshtype.hpp:975: error: expected `)' before '&' token
> 

I don't remember what the problem was for this one: just do a "diff" on
the Netgen directory between gmsh 1.60.1 and gmsh 1.64.1 to see what the
differences are.


> 
> Thank you again for your help.
> 
> Romain Quey
> 



-- 
Christophe Geuzaine
Assistant Professor, Case Western Reserve University, Mathematics
http://www.case.edu/artsci/math/geuzaine