Windows gmsh

[Christophe Geuzaine]

> What do you think of this?
> http://www.llnl.gov/casc/Overture/
> 
> or any of these meshers?
> http://www-users.informatik.rwth-aachen.de/~roberts/software.html

I don't use any of these software packages. I once used 'triangle' for
2D meshes, and it worked well.

> 
> To me, the appeal of GetDP is twofold.  (1) Design based on differential
> geometry.  (2) "Generic" nature of the solution library.
> 
> But as far as gmsh itself, what makes it "stand out from the crowd"?  The

Nothing (except that it's 3D and free) : it is mainly a secondary
project (in comparison with GetDP), used for testing purposes only. (it
is quite handy for postprocessing, though).

> ability to define a characteristic length is important, but many other meshers
> claim to produce an "optimum" mesh according to certain metrics.
> 
> Also, I have a hard time really understanding the correlation between the
> differential forms described at
> http://www.montefiore.ulg.ac.be/services/elap/elm/general_en.html
> and the actual triangular cells used by gmsh.  This article is excellent however
> and thanks for putting it on the web.

All the "finite elements" we use are based on differential forms (the
so-called "Whitney forms" and their generalization to higher degrees and
other geometrical shapes).

-- 
Christophe Geuzaine

Tel: +32-(0)4-366.37.10    mailto:Christophe.Geuzaine at ulg.ac.be
Fax: +32-(0)4-366.29.10    http://www.geuz.org