Google Summer of Code Application 2014

http://www.google-melange.com/gsoc/homepage/google/gsoc2014

Application Deadline

Status: Applied with the text below

Timeline: http://www.google-melange.com/gsoc/events/google/gsoc2014

• Friday, February 14: 19:00 UTC

Mentoring organization application deadline.

• Monday, February 24: 19:00 UTC

List of accepted mentoring organizations published on the Google Summer of Code 2014 site.

Old

• Base on 2007, 2008, 2009, 2010, 2011, 2012, and Google Summer of Code Application 2013

New

For discussion

Possible additions to last year's (2013)

[provisional]

proposed additional question for the application:

• HB: not really a question, but I propose we should reinforce that all questions must be answered by the students, they can't pick and choose the ones that make them look good. e.g. the student must be aware that if they fail to e.g. supply a timeline their application is incomplete and they won't be chosen.
• AG: I agree, and the text you added on top of application template looks good to me.

• HB: in the timeline question I propose we emphasize midterm deliverable(s). Also, what do you think about the idea of asking for internal 25% and 75% deliverables? (to avoid major shocks at the midterm and concertina workloads), with a note that 50,75,100% deliverables can be adjusted at quarterly delivery dates?
• AG: It would help, but I don't know how easy it is for the student to provide them before he/she starts coding. Maybe we can ask them to accepted students, the first week(s) after acceptance.

• HB: also re the all important timeline, I suggest we make it clear (somewhere) early on that learning how to compile, install, and use the software, and also spending weeks reading books is really stuff they should be doing before applying and on into the community bonding period. The previous years where coding for some projects didn't start until week 3-4 as the student didn't have the project code building yet isn't good enough IMO.
• AG: Right. Let's state clearly that community bonding period is meant to catch up with these things, and then coding period can start right away.

Possible removals vs. last year's

• Notably the student-application template has been removed from the application.

The Application

Organization id

osgeo

Organization name

OSGeo - Open Source Geospatial Foundation

Organization description

OSGeo is a non-profit organization serving as an umbrella organization for the Open Source Geospatial community in general and 33 code projects in particular:

Web Mapping: deegree, geomajas, GeoMoose, GeoServer, Mapbender, MapBuilder, MapFish, MapGuide Open Source, MapServer, OpenLayers, ZOO-Project, Team Engine

Desktop Applications: GRASS GIS, Quantum GIS, gvSIG, Opticks, KDE Marble

Geospatial Libraries: FDO, GDAL/OGR, GEOS, GeoTools, MetaCRS*, OSSIM, PostGIS, rasdaman

The following are sub-projects of MetaCRS:

• PROJ.4, GeoTIFF/libgeotiff, CS-Map, Proj4J, Proj4js, SpatialReference.org

Metadata Catalogs: GeoNetwork, pycsw

Other (non-code) Projects: Public Geospatial Data, Education and Curriculum, Live Handout DVD

We host the annual FOSS4G conferences (this year two of them) with typical attendance of 500-1000+ geospatial developers, industry and government types, and researchers. Our mailing lists collectively go out to ~ 20,000 unique subscribers.

For GSoC we hope to act as an umbrella org for a number of other non-OSGeo FOSS geo projects, including OpenStreetMap, pgRouting, PyWPS, istSOS, and uDig.

Organization home page url

http://www.osgeo.org

Main organization license

• Each member project uses their own, but all OSGeo projects are required to use an OSI approved license. (see http://www.osgeo.org/incubator/process/evaluation.html)
• Member projects typically use: GPL >=2, LGPL, MIT/X, etc.

GSoC program requires an OSI approved license, but the dropdown only lets you select one.

GNU General Public License

Veteran/New

Veteran

Backup admin

The username of the user who will serve as the backup admin for this organization.

• Hamish Bowman (melange: hamish)
• Dustan Adkins (melange: dadkins_mentor)

→ but I guess we just have to pick one melange ID here?

Carol wrote on the ML: "You can set one backup admin as part of your application in the "backup admin" field. Should you be accepted into the program, yourself and the person listed in that field will immediately be admins for your org and you can then add as many additional admins for your org at that time."

hamish

(or dadkins_mentor, but I'm guessing that my (HB) UID has a much longer track record in the system and that might be what they're looking for)

Why is your organization applying to participate in GSoC 2014? What do you hope to gain by participating?

(required)

We hope to get more students excited about both open source and geospatial development. This will help to grow the development community and of course tangibly advance particular projects. Additionally by way of the level of exposure that SoC provides we aim to further extend the use, development, and teaching & training of open source GIS at universities worldwide. This will help to prepare a new generation of developers for industry and government agencies capable of introducing and implementing viable open source geospatial technology, to the benefit of all.

Specifically, one of our main goals is to attract and nurture student participation in our project development teams and prepare a new generation of contributors to the rapidly growing and exciting geospatial field. As our past GSoC students become todays mentors, continuing participation in GSoC allows us to combine experience with fresh minds and bring innovation and experimental approaches to the projects.

We also hope to use this opportunity to bring open source geospatial projects closer and to communicate more. We are encouraging students to come up with ideas that cross project boundaries, and to find new ways to connect the separate efforts in order to bring out new and fundamentally improved wholes. After a previous SoC we saw the birth of a new project, Spatialytics, which is a novel combination of several geospatial technologies. This is, as stated above, one of our goals, which GSoC helps us achieve in a way which cannot be matched in any other regular effort.

Last but not least it has been a great opportunity to mingle code with and get to know other like-minded FOSS projects in our app domain. (such as OpenStreetMap, Mapnik, Sahana, Ushahidi, ...)

How many potential mentors do you have for this year's program? What criteria did you use to select them?

(required)

We have so far responses from potentially 15 to 30 mentors across OSGeo projects. Each OSGeo project chooses their mentors (two per student) according to past mentoring experience, involvement in core development, specific knowledge for the idea they wish to mentor. More people will show up as the ideas pages will grow. Last year we had approximately 50, and as GSoC is popular with our member projects we'd expect about the same level of interest again.

This year we have called on our newly formed ICA-OSGeo Labs network to participate to OSGeo GSoC. Together with the International Cartographic Association this network connects labs at 65 educational institutions around the world who primarily use OSGeo and open source software in general. These labs are spreading the word amongst their students and we look forward to see professors and lecturers take part as mentors.

http://wiki.osgeo.org/wiki/Edu_current_initiatives

What is your plan for dealing with disappearing students?

(required)

In order to minimize the chance of disappearing students we will ensure that the OSGeo administrative contact and the mentors have full contact information (email, home phone, etc) so that we can keep in touch with them at all times. We also stress to our students that this is a Job, as opposed to a school project, and require a timeline (including exams and vacations to work around) as part of the application.

In addition, we will make our reporting requirements clear up front. Letting the students know that we will require a weekly progress report to be posted to OSGeo's SoC mailing list (cc the member project's development mailing list), participation in public IRC channels and project mailing lists will all help ensure ongoing commitment to the work. Mentors are required to make sure that happens and that the lines of communication are kept open from beginning to end. Weekly progress reports are reviewed by all, and timeliness monitored and prodded by the org admins.

Other possibilities include:

1. Having potential students work with their potential mentor to prepare a plan for the implementation of their project.
2. Setting milestones that need to be reached by the student. These milestones would be defined before the student's work begins. A student would subsequently write a brief e-mail informing the OSGeo SoC mailing list (cc project-specific mailing list) when they have reached each milestone, giving a brief description of the work completed to that date. This is in addition to the weekly progress reports which are intended to be more casual and actively soliciting feedback.
3. Selecting students that have shown previous interest in the geospatial field and free software. A student who already has a (reviewable) FOSS track record has a much higher chance of success.

What is your plan for dealing with disappearing mentors?

Project steering committees will appoint replacement mentors if existing mentors become unavailable or unresponsive during the project period. We also have backup mentors for all projects so the backup mentor can fill in for, or replace the primary mentor in case of changing circumstances. Our administrators take an active role communicating with students and mentors, and so will be able to help out early if there is a problem somewhere.

Among all the OSGeo mailing lists there are 20,000 subscribers, indicating a good pool of people to draw from if required. Due to relatively high member project code and personnel cross-over, in a pinch a mentor from another OSGeo project can usefully help out. Additionally this year we plan on communicating with mentors even more closely thus if something happens with the mentor it should not come as a surprise and additionally there is a backup mentor to continue the mentoring.

What steps will you take to encourage students to interact with your project's community before and during the program?

(required)

The steps taken depend on the individual projects. As an example, in previous years we required weekly reports and encouraged the students to interact with us on the project mailing lists and IRC channels. We also encourage the developers and users who are not mentors to provide additional advice and feedback to the participating student and help with testing the students contributions during and after the program. In general other developers have not been shy about sharing their opinions, guidance, and advice on the member project development lists.

Furthermore the mentors are encouraged to actively involve the students in the community by having the student interact with the mentor through the community (public) channels rather then one-on-one. This also assists with student integration into the project and alleviates the disappearing-mentor problem.

We encourage projects and students to get together for in-person meet-ups during our large omnibus FOSS4G conferences, and at smaller project conferences and code-sprints. For example, in past years we had good success with this with students introduced to the development community and talking about their projects early in the summer: http://grasswiki.osgeo.org/wiki/GRASS_Community_Sprint_Prague_2011 and http://grasswiki.osgeo.org/wiki/GRASS_Community_Sprint_Prague_2012 .

In late March we will host a 5 day multi-project code sprint in Vienna, which unfortuantely is too early for interacting with students, but will let mentors get together and plan. Our 10th annual FOSS4G conference will be held in Portland, Oregon in September, and we would hope as many students and mentors as possible will be able to attend to present their work.

What will you do to encourage that your accepted students stick with the project after Google Summer of Code concludes?

• Socially it's a nice place to be: As a large and mature FOSS organization with a large academic presence we have had a lot of success fostering a professional and friendly community environment.
• Recognition: It is important to us that students feel they are part of the project and see their hard work appreciated.
• Results: By encouraging ongoing community engagement during the student-mentor period we hope to knit the student into the project over the summer, to avoid the situation of having the student-mentor interaction happening in private and the code branch languishing without buy-in from the greater dev community.
• Clear expectations: This year we are again asking the student to discuss the time after the summer in their initial application.

Are you a new organization who has a Googler or other organization to vouch for you? If so, please list their name(s) here.

n/a

Are you an established or larger organization who would like to vouch for a new organization applying this year? If so, please list their name(s) here.

n/a

If you chose "veteran" in the dropdown above, please summarize your involvement and the successes and challenges of your participation. Please also list your pass/fail rate for each year.

2007: 13/19
2008: 15/19
2009: 17/20
2010: 7/10
2011: 19/21
2012: 20/22
2013: 21/22

OSGeo participated in 2013, with 21 of 22 students successfully completing their projects. It was a great year for the variety of software projects which took part under the OSGeo umbrella. We were glad that a bunch of small, newly formed geospatial projects joined our GSoC as guests and established links with the established OSGeo network, as well as to have more established projects like OpenStreetMap join us. We sent two delegates to the mentors' summit, who lead sessions on managing GSoC umbrella orgs and FOSS mapping.

GSoC 2013 final report: http://www.osgeo.org/node/1410

In 2012, 20 of 22 students completing the summer. We provided GSoC-specific umbrella support for two smaller projects from outside of our Foundation.

Successes: All in all our students were great and a lot of good work was completed -- perhaps the least remarkable-sounding but most important success we had is how smoothly everything ran in spite of handling our largest number of students yet. Many of our former students continue to be involved in our contributing projects' development and take on roles as mentors. A couple of the smaller external projects we had taken under our GSoC umbrella in the past are now participating in our incubation process to become full members of our Foundation.

Challenges: Getting everyone to fill in their evaluations on time is always a challenge, but we're pleased to report that in 2013 we didn't have to resort to 3 AM phone calls to the other side of the world as everyone got theirs in on time, no small effort for an umbrella org! Unfortunately we had to fail one student at midterm, we took him on as a risk and had high hopes, but a combination of personality mismatch with his mentor, backup mentor and org admins not picking up on that early enough, and not a strong enough coding background lead to many diversions from the established timeline and only a few lines of useful code once the boilerplate commits were removed. This was especially disappointing as he was a friendly student who was quite enthusiastic about the program. We considered salvage with re-made timeline and goals and better mentoring support, but in the end despite promises there was no last minute code commit, and we had to judge on what little code we had in hand. All our students who passed midterm also continued on to pass the final.

In 2012 former GSoC student and mentor Anne G. smoothly transitioned into the role of lead admin from Wolf B., with Wolf and another long-serving co-admin standing by with advice. Since our umbrella is wide, we brought in an additional co-admin for better communication with mentors and projects that the other admins didn't know as well. Anne is a great communicator and has settled in well. (n.b. that wasn't written by her ;-)

OSGeo sent two delegates to the 2012 Mentor Summit, who chaired summit sessions on umbrella org admin'ing, geospatial FOSS, and humanitarian FOSS. We spent a lot of time comparing notes and ideas with other umbrella orgs and participating in Open Science sessions.

OSGeo participated in 2011, with a very good success rate. Many of the students continued to work with their OSGeo projects after the end of summer, and even become official contributors. The mentoring has been overall effective, and the communication among students, mentors and admins has been constant. Still, getting in contact with elusive participants represented the main difficulty, and has been addressed case to case. OSGeo sent 2 delegates to the 2011 Mentor Summit, and benefited of the presence of 2 more, who were already in Mountain View. As in the previous year, the delegates chaired a Geospatial session, and improved connections with many other projects.

OSGeo participated in 2010, and in many ways it was a tough year, but we like to think that we managed to come out victorious. It was challenging because our slot count was cut to about half of what we have had the previous year, while still having the same number, if not more projects participate under our umbrella. We also had some surprising student dropouts (one who is still MIA, as far as we know), maybe the aliens targeted OSGeo, because of our knowledge of the earth? ;) But despite these challenges we managed to put out some great projects and were in general very happy with the year. OSGeo sent 2 delegates to the 2010 Mentor summit who participated in many geospatial sessions and also chaired one session. We made new friends among the FLOSS geo-related projects and re-connected with old friends.

OSGeo also participated in GSoC in 2009. Twelve OSGeo member projects participated and were very successful. The applications were of better quality so we had fewer dropouts than the previous year. We also had further cross-project co-operation projects, which went very well indeed. OSGeo Sent two delegates to the Mentor Summit.

OSGeo participated in GSoC in 2008. Eleven OSGeo member projects participated and were in general very successful. OSGeo didn't send delegates to the Mentor Summit due to the busy schedules of the mentors. See http://wiki.osgeo.org/wiki/SoC_Report_2008 for a full report.

OSGeo participated in GSoC in 2007. Six OSGeo projects participated and were generally very successful. OSGeo also sent one delegate to the Mentor Summit in 2007. See http://wiki.osgeo.org/wiki/SoC_Report_2007 for a full report. We also opened up a demo theatre track highlighting SoC work at our FOSS4G conference that year.

OSGeo was formed in early 2006 and did not participate in GSoC prior to 2007. However, Refractions Research did participate in 2006 on behalf of the GeoTools, uDig, and PostGIS projects which now participate through OSGeo. Even after all this time most of the the previously participating mentors and administrators continue to be involved in this year's OSGeo GSoC effort.

[Back to Google Summer of Code 2014 @ OSGeo]