This is why we should release v3.4.3 and perhaps v3.4.4.  It takes time
after a PETSc release for downstream projects to update and make
releases.  This should in no way discourage or defer releasing v3.5, but
it gives a time scale for what is readily available to users.

--- Begin Message ---
On 2013-10-09 08:43, Corrado Maurini wrote:
> I would like to propose minor modifications to PETScSNESSOLVER.
> 
> Before proceeding I have two questions:
> 
> - Is it fine for everybody if I remove the support to PETSC 3.2? It
> complicates a lot the code and now PETSC 3.2 is obsolete (current
> version 3.4). Moreover SNES solvers for bound constrained problems are
> not reliable in 3.2.
> 

I would like to remove support for PETSc 3.2, but 3.2 is the only 
version available on Ubuntu. Version 3.4 has appeared in Debian, so will 
probably make it into Ubuntu next year.

If you work on the code, as a compromise I would suggest not supporting 
PETSc 3.2 in the SNES interface. As soon as PETSc 3.4 is widely 
available as a package, we can drop support for 3.2 and 3.3 throughout 
DOLFIN.

Garth

> - Currently the function PETScSNESSolver::FormJacobian uses an
> inefficient MatCopy of matrices after assembling, instead of passing
> them as reference. Neither me nor Patrick were able to find a way do
> it properly. May someone expert in shared pointers and friends help on
> this point? I reported the details in issue #132 on bitbucket
> (https://bitbucket.org/fenics-project/dolfin/issue/132/ [1]
> 
> Best,
> 
> Corrado
> 
> Links:
> ------
> [1] https://bitbucket.org/fenics-project/dolfin/issue/132/
> 
> _______________________________________________
> fenics-support mailing list
> [email protected]
> http://fenicsproject.org/mailman/listinfo/fenics-support
_______________________________________________
fenics mailing list
[email protected]
http://fenicsproject.org/mailman/listinfo/fenics

--- End Message ---

Attachment: pgp_q_pqJNYKS.pgp
Description: PGP signature

Reply via email to