GSOC

[Andy Ray Terrel]

Would there be interest in putting together a NumFOCUS Google Summer
of Code application? Its always good to have more seats for people
working on science codes.

-- Andy

[Jarrod Millman]

When I asked Google (a few years ago, granted) if it made sense for
SciPy to apply as a mentor organization, they voiced a strong
preference for us to remain under the PSF umbrella. The main concern
was that they wanted a lot of support from the mentor organizations
and the PSF was one that they felt did an excellent job. Also in past
years (when I was involved more), we often had trouble finding
willing mentors and qualified applicants. That may be different now.
My feeling would be to focus on getting our own program up and running
quickly, rather than splitting our efforts by becoming a GSoC mentor
organization.

I don't have a strong preference and my information is slightly dated,
so take my comments with a grain of salt.

Best,
Jarrod

[Fernando Perez]

I'm -1 on numfocus going in this direction, b/c I think the gsoc
program should be very close to the code, and therefore should be
managed by each individual project. The PSF has Python itself as its
core project, but in our case, each project has its own development
dynamics.

For example, IPython's gsoc experiences haven't been the best (no
fault of the students, mostly my own), while projects like sympy and
sklearn have had some great success on that front. It all comes down
to the specifics of each project, so I'm not sure that NF could really
help all that much.

I think our (limited) resources would be better spent on some of the
other fronts in which we're engaging.

Just my 1e-2, I don't want to sound too negative but it's my take on
the matter from experience.

Cheers,

f

On Wed, Mar 20, 2013 at 1:33 PM, Andy Ray Terrel <andy....@gmail.com> wrote:

> --
> You received this message because you are subscribed to the Google Groups "NumFOCUS" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to numfocus+u...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Travis Oliphant]

I personally think we should spend the time instead following-up with the other funding sources we have been put in the middle of and push our own Technical Fellowship approach which is broader and can close the "mentoring gap" where GSOC program doesn't help as much.

-Travis

[Ralf Gommers]

On Wed, Mar 20, 2013 at 10:51 PM, Jarrod Millman <jarrod....@gmail.com> wrote:

On Wed, Mar 20, 2013 at 1:33 PM, Andy Ray Terrel <andy....@gmail.com> wrote:
> Would there be interest in putting together a NumFOCUS Google Summer
> of Code application?  Its always good to have more seats for people
> working on science codes.

When I asked Google (a few years ago, granted) if it made sense for
SciPy to apply as a mentor organization, they voiced a strong
preference for us to remain under the PSF umbrella.  The main concern
was that they wanted a lot of support from the mentor organizations
and the PSF was one that they felt did an excellent job.  Also in past
years  (when I was involved more), we often had trouble finding
willing mentors and qualified applicants.  That may be different now.
My feeling would be to focus on getting our own program up and running
quickly, rather than splitting our efforts by becoming a GSoC mentor
organization.

I agree. The PSF is doing a very good job and the fraction of students that get accepted by the PSF to work on SciPy Stack related projects is quite high.

Ralf

I don't have a strong preference and my information is slightly dated,
so take my comments with a grain of salt.

Best,
Jarrod

[Andy Ray Terrel]

Thanks everyone for the thoughts.

I personally would like to see more scientific coding projects (not
SciPy/Numpy but things like FEniCS, PETSc, Elemental and so on), but
sounds like NumFOCUS isn't the right place to lead that. I was
thinking the role would be more, help manage the paperwork and have
people using these projects have some money to get summer help.

-- Andy

[Ralf Gommers]

On Thu, Mar 21, 2013 at 8:39 PM, Andy Ray Terrel <andy....@gmail.com> wrote:

Thanks everyone for the thoughts.

I personally would like to see more scientific coding projects (not
SciPy/Numpy but things like FEniCS, PETSc, Elemental and so on),

I think the PSF would be happy to see those projects apply under its umbrella (assuming those projects have >= 3 mentors).

Ralf