[petsc-dev] not everything belongs in PETSc; thread/task management etc

[Matthew Knepley]

On Sat, Sep 27, 2014 at 12:16 AM, Barry Smith <bsmith at mcs.anl.gov> wrote:

>
> On Sep 27, 2014, at 12:10 AM, Jed Brown <jed at jedbrown.org> wrote:
>
> > Barry Smith <bsmith at mcs.anl.gov> writes:
> >
> >>   Perhaps we erred in keeping the thread/task management/model code
> >>   in PETSc as opposed to a small portable library? We should revisit
> >>   this, but unfortunately cannot before next Wednesday :-), or can we
> >>   :-)
> >
> > I was at the BLIS workshop yesterday and today discussing this with
> > Field and a guy at Q-Chem.  I think an independent library is desirable
> > in the long term, but I don't think it's something that we can "win" in
> > the sense of having everyone adopt our library.  So we'll always need to
> > interoperate with other systems.
>
>   Sure, but it will be easier to interoperate with other systems if it is
> a small, non-PETSc dependent beast.


My questions would be:

 - does it help development if it is independent

 - does it help adoption if it is independent

I am not sure the answer to either of these is yes.

  Matt


>
>   Barry
>
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20140927/4d3a2bda/attachment.html>