ffc team mailing list archive
-
ffc team
-
Mailing list archive
-
Message #03554
Re: [noreply@xxxxxxxxxxxxx: [Branch ~ffc-core/ffc/dev] Rev 1601: Comment on domain2dim and entities_per_dim in fiatinterface.py.]
On Sat, Jan 30, 2010 at 08:08:26PM +0100, Kristian Oelgaard wrote:
>
>
> On 30 January 2010 19:51, Anders Logg <logg@xxxxxxxxx> wrote:
> >It would be good to collect all these in one place, but UFL cannot
> >depend on UFC and UFC cannot depend on UFL.
> >
> >Are they needed in UFC? If not, it would be natural to collect them in
> >UFL. Then they can be used by both FFC and UFL.
>
> I don't think that we use it in UFC, but we do work with UFCInterval, UFCTriangle etc. in FFC and DOLFIN.
> Let's put it in UFL, it doesn't matter to UFC how the internal representation is as long as the output of "interval" is a cell defined by the vertices [0, 1].
>
> Kristian
Sounds good. I have added a FIXME about it in the code. We can do
another round of cleanups for 0.9.1 and fix it then.
--
Anders
Attachment:
signature.asc
Description: Digital signature
References