On Tue, Jun 25, 2019 at 9:39 AM Sage Sharp <[email protected]> wrote: > > On Mon, Jun 24, 2019, 9:08 PM Roman Shaposhnik <[email protected]> wrote: > > > On Mon, Jun 24, 2019 at 6:34 PM Sam Ruby <[email protected]> wrote: > > > > > > On Mon, Jun 24, 2019 at 9:18 PM Sage Sharp <[email protected]> > > wrote: > > > > > > > > On Mon, Jun 24, 2019, 4:09 PM Roman Shaposhnik <[email protected]> > > wrote: > > > > > > > > > On Mon, Jun 24, 2019 at 4:07 PM Griselda Cuevas > > <[email protected]> > > > > > wrote: > > > > > Question for Roman and other ASF Directors: is it fair to assume that > > > from a selection point of view, Outreachy is sufficiently similar to > > > GSoC, so something along the following lines would suffice: > > > > > > Would you be comfortable if the D&I committee were to adopt/adapt the > > > GSoC process? > > > > > > http://community.apache.org/mentee-ranking-process.html > > > > > > If not, what additional constraints need to be met? > > > > To me that's not a *selection* that's a *vetting* criteria. Sure, I > > trust Outreachy in that > > as much as I trust GSoC. The question is still what do we do if we > > don't have enough > > interns. > > > > The only plausible way out of it would be to let interns themselves > > deicide which TLP to contribute to (without any influence of ASF or > > D&I whatsover). > > > > Is this what is being proposed here? > > > > That is what I was proposing. Any ASF community could submit a project > proposal under the ASF Outreachy community listing. The project list could > include competing projects. There could be more projects listed than the > ASF (or sponsoring companies) have committed to fund. > > During the contribution period, the Outreachy applicants choose which > projects they want to contribute to. The applicants choose which projects > to contribute to after looking at the project listing on the Outreachy > website. There would be no influence from the ASF or D&I board. > > Projects which are ready for newcomers and that have fewer barriers for > people from marginalized groups will naturally end up with more applicants > that have completed a contribution. Projects that aren't will have less > applicants. That can trigger a review by the ASF D&I committee of what > barriers people faced in projects that didn't have as many applicants. > > ASF projects will learn a lot about the barriers people face simply by > participating in the contribution period. Outreachy applicants will get > stuck, and will reach out to mentors for help. What questions they ask, > where they get stuck, is documentation of what barriers people face. For > example, a couple rounds ago one free software community discovered that it > was impossible to compile a development environment on 5-10 year old > laptops. They didn't have any applicants who completed a contribution, but > that was valuable knowledge that they took to lower the barriers to entry > to their project. > > Sometimes mentors find themselves with not enough time during the > contribution period. That's probably a sign that they don't have time for > an intern. The end result is that applicants try to make a contribution, > get frustrated by the lack of response, and then pick a different project. > It's not great for the applicants, which is why we try to make mentors > aware of the time commitments up front. > > Basically, the contribution period allows applicants to chose which > projects are suited for accepting an intern. > > There still needs to be a set of coordinators who decide whether to list a > project in the first place. That criteria should be applied in the same way > across all proposed projects, regardless of whether the project is > competing. > > IMO, a project should be listed if: > > - The project will not disappear in 3 months. There's a long time > between when project submissions are due (February 26 or September 24) and > when the internship starts (May 20 or December 3). The project should still > be valid to be worked on at the end of that period. That means submitted > projects should not be time-critical for the community. > - The project has a set of clearly defined goals/milestones for the > internship. The project milestones should take into account that the intern > will spend 1-2 weeks ramping up on the project. > - Project mentors should have defined a set of newcomer-friendly issues > for applicants to work on. Projects should prepare to have 5-10 applicants > who work on 1-3 contributions each. There should be easy contributions to > test whether the applicant can set up their contribution environment. There > should be medium-sized contributions that will test the applicant's skill > levels. > - The project should have good documentation for how to make your first > contribution. Ideally projects would have a newcomer test their > documentation before the contribution period opens. > - The project mentor can commit to the 5-10 hours during the 5 week > contribution period, and 5 hours per week during the 14 week internship > period. > > Projects with two or more mentors should be given priority, since they're > more likely to be responsive to applicants and contributors. Projects with > mentors who aren't new mentors should be given priority. If a mentor has > participated in Google Summer of Code or another mentoring program like > Rails Girls Summer of Code, the project should be given priority. > > The goal for the above criteria is to list projects which have a high > chance of being able to accept an Outreachy intern. It's fair, and allows > for competing projects to be listed. > > TLDR: The Outreachy contribution period naturally allows projects which > will better support interns rise to the top. Projects that aren't ready for > interns will still learn where the barriers to entry are. The ASF can let > applicants make the choice of which project to work on by listing competing > projects under their community.
Thank you. That clarifies the proposal perfectly. Thanks, Roman. P.S. It feels like this is the closest (and cleanest) we can come to take care of the neutrality concerns. I also feel that we will still have to tackle "ASF now pays for software development" optics of this and make a sort of executive decision around whether the pros of the program outweigh the cons of the blowback. Or not.
