dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

On Fri, Dec 08, 2006 at 10:10:14AM +0100, Johan Jansson wrote:
> On Thu, Dec 07, 2006 at 10:26:33PM +0100, Garth N. Wells wrote:
> > 
> > Sorry, I see that no build system changes were made. I would still 
> > prefer all non-bug related changes to be make in dolfin, which will 
> > eventually propagate through to dolfin-stable.
> > 
> > It won't be possible to keep modules stable when making changes to the 
> > kernel. Interface in the kernel changes will affect modules.
> > 
> > Garth
> > 
> 
> This is fine though. Of course you have to propagate the changes to
> the kernel so that the modules still work. But you should not add new
> functionality to the modules or add new modules.
> 
>   Johan

What do you mean? Do you mean that development of modules should not
take place in DOLFIN, but instead in dolfin-stable?

This sounds a bit strange.

The purpose of dolfin-stable must be to have a stable branch with bug
fixes that anyone who is not heavily involved in the development of
DOLFIN can work against in between stable releases. This can include
graduate/msc students at KTH and Delft, users of the DOLFIN mesh at
Simula and everyone else.

/Anders