Google Summer of Code/2015/Organization application
https://www.gentoo.org/main/en/irc.xml (#gentoo-soc on freenode)
Why is your organization applying to participate in Google Summer of Code 2015? What do you hope to gain by participating?
We intend to recruit enthusiastic, experienced, high-quality Gentoo developers to maintain the growth and health of our developer community and its many derivative distributions, including Google's ChromeOS. We also contribute on a regular basis as an umbrella to entirely separate OSS projects in ways that benefit Gentoo — PackageKit is one excellent example.
How many potential mentors do you have for this year's program? What criteria did you use to select them?
Our top-ranked projects are largely student-initiated ideas, so we pull in the best-suited mentors from our community of more than 200 active Gentoo developers on a ad-hoc basis to work with each student. We also have a pool of around 20 mentors who are involved in evaluation every year and are happy to mentor relevant projects. Project-specific mentors will be selected if they are known as the authority on that project in Gentoo in addition to having the attributes in our mentoring guide <http://www.gentoo.org/proj/en/userrel/soc/mentoring.xml>. Senior overseeing mentors with previous mentoring experience will also be attached to teach project mentors. To ensure that mentors can do a great job, we need to also have high-quality project administrators. Admins must have past mentoring experience, and at least one admin must have past experience doing so.
What is your plan for dealing with disappearing students?
If a scheduled meeting is missed, we will begin daily attempts to reach them through multiple forms of communication. We will require a physical address and phone number, and we will confirm the phone number at the start of the program. If we cannot reach them successfully via these methods, Gentoo's community is large, so we will attempt to have someone near them geographically get in touch. If we cannot get in touch for 1 week without any advance notice from the student, they will be sent a final warning. If we hear nothing by the following day, they will be failed. Students will be informed of this policy when the program starts and will agree to follow it.
What is your plan for dealing with disappearing mentors?
We have available backup mentors so that students will never be without mentorship, even if the original mentor disappears. We will pursue the same communication methods as with students, and failures will be dealt with by blocking them from any future mentoring.
What steps will you take to encourage students to interact with your project's community before and during the program?
We will strongly encourage applicants to interact with the community using our standard communication methods (mailing lists and IRC) before and during the application & evaluation periods. In fact, this will be part of our custom application template. If they cannot learn to do it during that month-long period, we can't expect that they will learn to do so during the next few months. That will count against them in the ranking of their application. Since communication will be one of the requirements for a successful application, we expect that problems during and after the program will be much rarer. We will treat students in the same way we treat other new members of our community, a significant portion of whom are college students just like the applicants. By encouraging students to communicate directly with the community instead of privately with their mentors, we will infuse them with the process of open-source development. In the past, we've discovered that a major key to sticky contributors is close personal ties with the community. Requiring students to become full members of the development community instead of segregating them with their mentors as conduits to the community will naturally result in a higher stickiness. Another technique that we discussed ever since the 2008 mentor summit was converting students into developers before the summer ends, to create a natural continuation instead of a break. In 2009, we proved these techniques work by increasing our new-developer ratio from 20% to 67% of passing students. In 2008, we had 20% become developers but 60% stay involved in the community. In 2009, we successfully converted the community members into developers instead of losing their potential contributions. In 2010, we grew the size of our program by more than 100% and proved that we could scale our successful methods up to a 19-student program by maintaining our high recruitment rate. In 2011, we succeeded at our goal of admitting and passing students with projects that were more relevant to Gentoo's core rather than peripheral side projects. We additionally continued with our high level of pass rates and recruitment.
What will you do to encourage your accepted students to stick with the project after Google Summer of Code concludes?
Donnie Berkholz wrote a detailed blog post on what we do here: http://redmonk.com/dberkholz/2012/07/10/how-to-recruit-open-source-contributors/
If you chose "veteran" in the organization profile dropdown, please summarize your involvement and the successes and challenges of your participation. Please also list your pass/fail rate for each year.
Gentoo has been involved with GSoC for the past 9 years. A total of 76/89 projects were successfully completed by our students. Some of our biggest challenges were in handling transitions between org admins, in part because it's so rare. We now have two primary admins. Previous transitions have been abrupt rather than gradual so we think this will be a great help. Another ongoing challenge is outreach, particularly as developers continue to move toward cloud-based technologies and as OS X becomes an increasingly popular alternative to Linux. To combat that, we've developed editions of Gentoo that run on OS X ("Prefix") and aim to integrate more tightly with the cloud through a variety of GSoC projects. We're also planning more extensive outreach efforts this year to reach the our best prospects. 2006: 10/14 2007: 8/9 2008: 5/6 2009: 6/7 2010: 16/19 2011: 14/15 2012: 8/8 2013: 6/7 2014: 3/4