My notes on the foundation idea
Fernando Perez
a few months back Travis and I had an email exchange about some of
these ideas, so I'm just going to paste hear for the list the reply I
sent out to him at the time, which had most of my views on the matter.
Cheers,
f
I may have mentioned this to you before, but I think a good model to
look into and learn from would be the Linux Foundation/OSDL, the guys
who actually pay Linus Torvald's salary. That setup provides a space
where big vendors who sell Linux-based products (say HP, IBM, Oracle)
can fund the part they have a common stake in, the Linux kernel
itself, despite being direct competitors out in the field. Given the
interest in numerical/scientific Python in the financial world, I
think a model like this could be attractive to some deep-pocketed
firms who could see a good cost/benefit ratio, without needing to
involve themselves directly with python itself. It may require some
selling of the idea, but I think it's actually sound and thus can be
presented in a convincing manner.
We should also talk to Chris DiBona, at google, who manages their open
source interactions, he might have some useful insights. And
obviously the PSF itself can be a source of information, but since
you're a member I figured that would have already been on your radar.
One important thing to try to find out about: how to get around the
overhead costs of most academic institutions, which are typically ~50%
(and can be even drastically higher at some places like DOE
facilities). Because it seems clear to me that one thing such a
foundation could offer limited funding to academics: Enthought's
funding of Brian and me for ipython, or William Stein's compmath grant
for multiple sage days sprints, are examples of the kind of targeted
support that could be done by this entity. But if such support is
provided as part of someone's regular academic work (to pay e.g. for
summer salary and some students), the universities will immediately
want to get their cut of overhead. I remember when Enthought gave me
some money to pay for Gael at Berkeley, we found ways to reduce that
overhead, but if I recall correctly we couldn't bring it down to zero.
Once concrete funds become available, this is something worth doing
some research into.
Of the list you mention, I agree totally on the first two items, and I
imagine that for the third you mean supporting the existing ones
rather than starting a new one? One thing I'd love to see would be
even more funding for students, but also some travel grants for
members of key projects even if they are not students, if they present
a credible proposal for a sprint. If a project has a few members who
can't make it to the conference but who are not students, and they
submit a proposal that includes solid sprinting, I think it would be
fantastic to be able to support them. I can easily imagine someone
like Michael Droetboom being more able to attend if he doesn't have to
find travel funding and instead needs to get only leave authorization,
just to name one example.
Another idea (I'm just brainstorming here, while waiting for my flight
at the Miami airport): contacting the Knight foundation. Those guys
have had software-related funding initiatives in the past, and they
could be a good source of ideas and information. At some point it
could even be possible to partner with another foundation such as them
to match funds and provide more resources, if a project is found that
can match the interests of both.
OK, I don't want to turn this into a novel. I hope these somewhat
random points have some value, and needless to say, this is *super*
high on my interest list, so I'll be delighted to continue the
discussion.