On Tue, Feb 19, 2013 at 4:09 PM, Aaron Meurer <[email protected]> wrote: > There are some issues with this, which would need to be resolved if we > were to do it: > > - First off, we need to determine if this is actually allowed under > the rules of the program. I know that other organizations, such as > the Python Software Foundation, act as umbrella organizations, but > there may be some caveats that we should be sure of. > > - Second, there would obviously need to be full support from the > secondary community. > > - Third, we need mentoring support. I would say that we'd need at > least one mentor who is a SymPy developer and one mentor who is a > developer of the secondary organization (they may end up benig the > same person, if someone works in both places). > > - Finally, I we would need to decide, as a community, if we want to do > it. I personally would want to evaluate this on a per-idea basis, so > I can't say if I would support it or not without more specifics. I am > not categorically opposed to the idea at any rate. My opinion isn't > the end all, though. So let's see how others feel too. > > - And of course, we would need an actual student who is interested in > doing this. > > One thing to consider here is that in the two years that we are > limited in the number of slots that Google gives us. It might be more > worthwhile to pursue such an idea as a project under the PSF, who > usually gets lots of slots, and uses most of them for umbrella > projects. Of course, we could also just ask the students to apply to > both, and we could apply the project to whatever organization ends up > with the slots (this is how things worked back when SymPy just > participated under the umbrella of other organizations; for example, I > submitted my proposals to PSF and PSU when I was a student, and I was > accepted under whichever organization had the slots for me). > > I definitely agree that more projects using SymPy as a dependency is a > good thing for SymPy. Not only does it indirectly give us more users, > but it allows us to see concrete ways that our code is being used > (assuming the library is open source), and forces us to make our code > more amenable to be used as an API (e.g., make classes friendly to be > subclassed, which they currently aren't, see issue 3652). This > doesn't just apply to statistics, though that's a great example.
I agree with everything you said Aaron. It would be good, but there are the caveats you mentioned. Ondrej -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sympy?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
