Using GSOC to train non-programming students

Participants: Bill Kendrick, Andrew Lake, Martyn Shaw, Charles Severance, Hilmar Lapp

Clarification: there are two kinds of "non-programming students":

Students most skilled at (and interested in) software engineering tasks that don't involve much or any actual coding, such as documentation, usability testing and improvement, UI and look-and-feel design, human-computer-interaction improvement.
Students who have some preliminary programming training (such as from university classes), but aren't experienced coders yet, though would like to acquire the ability, but won't find the time without some hand-holding from a mentor.

Proposal 1: allowing non-programming contributions

Google seems to almost prohibit training level applications
- The C-word
- Need to justify to Google that an idea 'qualifies'
- shouldn't the impact of a project for a project count?
documentation, usability, look&feel are a huge need in FOSS
- number of nerds typically isn't the limiting factor towards market share, there is a large talent pool to draw from
- accessibility features can also be a unique contribution
HCI school students or similarly strong credentials
- though these people may not stick with one project; however, they may stick with the community in general, taking experiences with them, and the likelihood of coders to stick around isn't that high either
would need good guidelines
would need capability to find 'non-nerd' projects

Students who want to jump into programming

needs a strongly technical mentor who is also willing to 'teach a class'
is more for a different kind of program
mentor ends up teaching rather than mentoring, teaching isn't going to happen much
however, there has been no pushback from Google if learning is part of the project
- judgment call of the mentor
- there are success stories, even if rare, and those can be empowering
student selection priority can be on staying on
such projects are high risk, organization and mentor needs to apply judgment
- high-risk projects have different criteria for what is expected than safe projects
- however, mentors need and can make the decision and have the necessary judgment

Proposal 2: have a 3rd evaluation option besides just pass and fail

option to terminate, with prorated payment
analogous to 'honorable discharge', 'didn't work out'
possible at any time, even early on
payment should be prorated, just as it is for 'normal' internships
would also encourage high-risk projects (which can be more innovative) and treating the results honestly from the beginning