On Mon, Aug 11, 2014 at 6:52 PM, Geoffrey Irving <[email protected]> wrote:

> Is there a reason why PetscDSSetDiscretization takes PetscObject
> instead of PetscFE?  The whole FEM API changed since I last looked at
> it, and having untyped routines makes it somewhat difficult to track
> the changes.


I did this to a) make it backwards compatible with what was there before and
b) because I was reluctant to start creating a big type hierarchy. This way
I can
have it take PetscFE or PetscFV and maybe something else. Then I figure out
where a type hierarchy makes sense, or a single type, or this just does not
work
at all.

   Matt


>
> Geoffrey
>



-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener

Reply via email to