[Dune] how to integrate curvilinear tetrahedral elements into DUNE/Alugrid ?

[Steffen Müthing]

Sorry for the empty post, my mail client is confused…

If the idea is to create an integrated solution, I think the idea Markus
brought up is probably the easiest solution: Just fork dune-{geometry,grid}
on GitHub and use those as a publicly visible repositories. Then you can
periodically merge in the upstream changes from our repositories and once
the new implementation works, we can just merge it back like any feature
branch. That way, if you are stuck and need help by a developer, we can just
check out those repositories and take a look (or simply browse them on the
GitHub site). That’s much easier than keeping the stuff on your local machines
and then having to send patches and stuff….

Pulling in the upstream changes is almost no work, I do it regularly for the
EXADUNE repositories. Last time Git forced me to resolve conflicts in less than
ten files for all core modules + PDELab, and the changes in those repositories
are much more intrusive than extending GeometryGrid and adding a new
Geometry implementation.

Just my 2c

Steffen

Am 23.01.2014 um 12:29 schrieb Christian Engwer <christian.engwer at uni-muenster.de>:

> On Thu, Jan 23, 2014 at 10:08:00AM +0000, Andreas Dedner wrote:
>> So you are suggestion a branch of dune-grid?
> 
> Not necessarily, actually I think a fork of GeometryGrid is better, as
> it allows to keep the current dune-grid. But I don't think that we
> want to (on the long run) to maintain two seperate grids which share
> most of their code, but we should try to merge them again.
> 
> Christian
> 
>> 
>> 
>> Dear all, while forking for the development is surely a good thing, why
>> would we want to rename this thing again. Basically the two
>> implementations will (hopefully) only differ in the choice of the
>> geometry implementation. I think we should try to share as much cde as
>> possible and thus try to merge the functionality into GeometryGrid,
>> ,,all'' that needs to be done is to provide a more general way of
>> specifying the transformation and storing its data. my 5 cents Christian
>> 
> 
> -- 
> Prof. Dr. Christian Engwer 
> Institut für Numerische und Angewandte Mathematik
> Fachbereich Mathematik und Informatik der Universität Münster
> Einsteinstrasse 62
> 48149 Münster
> 
> E-Mail	christian.engwer at uni-muenster.de
> Telefon	+49 251 83-35067
> FAX		+49 251 83-32729
> 
> _______________________________________________
> Dune mailing list
> Dune at dune-project.org
> http://lists.dune-project.org/mailman/listinfo/dune
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://lists.dune-project.org/pipermail/dune/attachments/20140123/621cd56b/attachment.sig>