On Wed, May 30, 2012 at 1:57 PM, John Peterson <jwpeterson@gmail.com> wrote:
One question about the design here:  H(div) elements will presumably
require a different type of mapping [0] than what FEBase currently
does with Lagrange.

So, should the geometry mapping stuff really be moved up into
FEAbstract, or would it be better to have virtual interfaces for that
which can be redefined by FEVectorBase?

My thinking was to create a PiolaTransformation object which augmented the current mapping within the FEVectorBase class because it looked like to me that you always have a composition with the H1 conforming case. But maybe I missed something? It also may be more clear to explicitly do this for each "branch" of FE. Thoughts?

Thanks for the comments.