[Dune] dune-geometry and Dune::Geometry

Oliver Sander sander at mi.fu-berlin.de
Thu Oct 13 17:39:45 CEST 2011


As an addendum to the discussion on FS 957 I should mention that I think
that it would be a cleaner design if BasicGeometry was independent 
(technically
and conceptually) of a grid.  I can't be too difficult technically, because
I have been using BasicGeometries without any grid on various occasions 
already.

best,
Oliver

Am 13.10.2011 16:03, schrieb Jö Fahlke:

> Hi!
>
> Christoph Grüninger has been working on dune-geometry for the last 1.5 weeks
> and its coming along pretty nice.  It isn't quite ready for general scrutiny
> yet -- we will tell you when we think its worth your time to look at it -- but
> there is one question that we have stumbled upon for which we need your input
> (see also FS#959[1]).
>
> [1] http://www.dune-project.org/flyspray/index.php?do=details&task_id=959
>
> This is the question of how much of Dune::Geometry to actually move to
> dune-geometry.  Dune::Geometry itself is an interface class for parts of a
> grid and as such takes the grid as a template parameter, so it definitely
> belongs into dune-grid.  In FS#957[2] there is a discussion about the
> internals.  From that I gather that GenericGeometry::Geometry,
> GenericGeometry::LocalGeometry and GenericGeometry::BasicGeometry belong into
> dune-grid.  But stuff like the mappings may go into dune-geometry.
>
> [2] http://www.dune-project.org/flyspray/index.php?do=details&task_id=957
>
> Is that about right?  Can you give us some hints on what should definitely go
> into dune-geometry and what should definitly stay in dune-grid?
>
> Thanks,
> Jö.
>
>
>
>
> _______________________________________________
> Dune mailing list
> Dune at dune-project.org
> http://lists.dune-project.org/mailman/listinfo/dune




More information about the Dune mailing list