[Gmsh] Local referential for the hexahedron and prism element family

[Geordie McBain]

On Mon, Feb 15, 2010 at 1:45 PM, Rui Maciel <rui.maciel at gmail.com> wrote:
> I've noticed that according to the manual's section on node ordering the local
> coordinate system referential for the hexahedron element family is positive
> while the one used for the prism family is negative.  Is this intentional or
> is it just a small bug on the prism's ASCII art representation?

How do you define the sign of a numbering of the vertices in a
polyhedron?  Is there a definition?

I know for tetrahedra there are two essentially distinct possible
numbering, so one could call one positive and one negative, but does
that generalize to polyhedra with more than four vertices?  I would
have thought that there would be more possible enumerations, since in
a tetrahedon each vertex is joined to all the others by edges, but
this isn't the case for triangular prisms or hexahedra.  (The
numbering of nodes in tetrahedra came up on this list in June
http://www.geuz.org/pipermail/gmsh/2009/004545.html
http://www.geuz.org/pipermail/gmsh/2009/004563.html.)