NDSA:Summer of code
NDSA Innovation Working Group
GSoC NDSA Innovation Call Notes
These are the notes from a conversation with Tim Donohue of DuraSpace who has been managing their Google Summer of Code program for several years.
Key information: The organizational application is actually short, mostly information about the organization itself. What they do want to see though is a URL for a public web page for the organization where there is a set of developmental ideas listed.
For reference, here are the DSpace and Fedora ideas pages from last year. https://wiki.duraspace.org/display/GSOC/DSpace+Summer+of+Code+Ideas https://wiki.duraspace.org/display/GSOC/Fedora+Summer+of+Code+Ideas
Any organization thinking about mentoring should probably start putting together this kind of ideas page.
Most successful projects
From DuraSpace’s , the projects that have been the most successful were the ones that were well scoped by the time that the students actually got down to doing the work.
The value of participating in Summer of Code
- The following were some suggested benefits for mentors for participating in GSoC
- Helps to attract new developers to the projects and community. In the case of DuraSpace, some of the GSoC participants have become dSpace or Fedora developers.
- Participating helped increase visibility of the projects to an audience that generally does not hear about them. It is mostly computer science students who participate here and being involved in GSoC exposes these students to the projects.
- There is a value to mentors, it is a useful experience for resumes but it also provides an opportunity to firm up ideas about the software.
- The low stakes on the projects make them valuable ways to do investigatory work. They become nice ways to get new ideas, different frames for thinking about the problems.
- If you’re organization hasn’t been doing open source software development for that long the act of thinking through the process of having someone remote who doesn’t have 3 months to absorb the details of your system is itself a valuable way to ease into practices for open source software development. That is, this can serve as a way to ease into and explore how you might build a development community around your organizations software development projects.
Tim’s Guidance on how the process for GSoC works:
- Each year, you (re-)apply to become a Mentoring Organization (if this is your first year applying, Google sometimes accepts recommendations/vouchers from orgs they have worked with in the past)
- Begin to brainstorm some potential student project ideas & start looking for possible mentors (as soon as you can)
- Assuming you are accepted by Google as a Mentoring Org, then you need to request a limited number of "project slots" from Google (this is essentially the maximum number of projects you hope to mentor that year)
- Based on demand, Google assigns you a maximum number of project slots (may or may not receive your requested number of slots)
- As a Mentoring Org, you then begin accepting student applications for those project slots.
- As a Mentoring Org, you get to decide which student applications to accept. You also are in charge of matching students with mentors and/or projects (and you can add new mentors/projects at last minute and students can also suggest project ideas to you)
- If any students are accepted by multiple Mentoring Orgs, then the student along with Mentoring Orgs need to decide which single project the student will work on. If you have extra project slots (ones that you couldn't find a student for), you often can give those to other mentoring organizations that can use them.
- Projects get finalized. Students start meeting mentors (virtually)
- Projects start & students begin working on project & meeting with mentors
- At mid-term, students get a pass/fail grade, assigned by their mentors. If failed, the project stops immediately & student gets $0. If passed, the project continues.
- At end of summer, students get another pass/fail grade (for second half of project), assigned by their mentors. Students also get an opportunity to grade their mentors & their mentoring organization.
- In fall, Google holds a "Mentor Summit" (on site at Google Headquarters). Each Mentoring Organization can send a maximum of two people to attend.
Below are notes from the NDSA Challenge Challenge Workshop on this tipic from, July 20, 2011. They can serve as the basis of planning for this group.
Google Summer of Code Workshop Notes
- Google Summer of Code proposal needs to be something that can be done in 6-8 weeks. What do you get out of it, and what do they get out of it? For this year, see: http://www.google-melange.com/gsoc/homepage/google/gsoc2011.
- To submit a project you either:
- specify problem or tasks, or
- you can provide content and ask “ what can you do with this?”
- We could propose two projects: one problem and one content set.
- Problem areas:
- Open source tool for data collection tool about standards and practices.
- Crowd-sourcing transcription tools – merge different approaches (structured data vs. just entry of text).
- Crowd-sourcing transcription of video – tag it in different ways
- APIs – Google data visualization plugins or data source plugins -- extending those tools for our community data. Need some way to select facets or elements of the metadata standard.
- Extract metadata using API -- for example from a geodatabase (Generalize extraction of metadata).
- Bag-it plug-ins for gmail.
- Quantify user feedback on what people want or are using, and the way people are using it (maybe including sitesearch analytics).
- Compelling ways of linking data sets using open linked data concepts or other concepts.
- Build something that is an ongoing marketing tool of NDSA. Point them to source materials and develop some cool apps (i.e., geo-located data).
- Wallpaper for NDSA image of the day.
- Digital preservation clock.
- Schedule for Google Summer of Code application (February-March). March 11 was the deadline for 2011, for example.
- Action Item #1: Kris and Carolyn will get together on proposal for crowd-sourcing for transcription options – possibly including linked data and other options (cite examples like NYPL).
- Action item #2: All will propose possible publicly available content sets that we might put together as a proposal for building a cool app.
- GML relay. Vendors import from each other and exchange data. For the purpose of pushing open standards.
- Hackfest (UK, usually regional). Pick set of challenges. Or THATCamp.
- Topcoder – coding contests.
NDSA Workshop Participants
- Kris Carpenter Negulescu
- Micah Altman
- Raphael Barbau
- Kate Murray
- Michael Levy
- Carolyn Sheffield
- Liz Madden
- Steve Morris