The projects are in and the mentors have filed their evaluations. In the final tally, it looks like Google’s Summer of Code 2006 was a success.
The program, having just finished its second year, provides funding to students and open source organizations for project development.
In 2005,
Google accepted 410 projects. In 2006, Google’s Summer of Code accepted
a whopping 630 projects spread across 102 mentoring organizations. In 2005, there were only 41
mentoring organizations.
Each student who successfully completed their
projects over the summer were to receive $4,500 from Google, while each open source organization was to receive $500 per successful student.
Leslie Hawthorn, open source program co-coordinator at Google, told
internetnews.com that the program had an 82 percent overall success
rate in 2006.
For those who didn’t complete the program,
Hawthorn didn’t see any one common thread.
“More often than not it was a ‘life happens’ situation when students drop
out,” Hawthorn said.
Hawthorn and Chris DiBona, open source program manger at Google, cited one particular student effort as being somewhat of a surprise.
Steffen Pingel was voted in as a committer for the Eclipse Mylar project as a result of his Summer of Code efforts.
“Think about it, this student went from being a college student who knew
nobody in the project to becoming a peer with people from everyone from IBM
to BEA to you name it in just a matter of a summer,” DiBona said.
“The whole point of Hawthorne expects one of the goals this year to be for students to stick with their projects. She actually also has a plan for improving retention in future years of the program. The idea is that if the student She noted that the program has a lot of support “I get to make more of them and help move projects forward and bring
Another interesting effort cited by DiBona yielded an OpenGL GUI Widget
system in Ruby
“Being able to do that within the framework of Ruby along with OpenGL so now
you can make a program in Ruby that can run on any platform that will give
you a great user interface is a cool idea,” DiBona said.
Google was also a mentoring organization for the Summer of Code,
though students didn’t necessarily work on projects directly for Google.
“The reason why we have the Google category is not for the things that we
want, which sounds kind of weird,” DiBona explained.
the Google category is that there are some students that are working on
things that are extremely specialized and can really only be judged
adequately by their student adviser at school, so we allow for a certain
amount of those kinds of students. It’s more procedural than it is strategic,
I guess.”
From a financial point of view, Google’s 2006 Summer of Code effort pumps
millions of dollars into the open source community. Yet there isn’t a
tangible return on investment (ROI) that is easy to measure.
“We get a benefit from every project is how I see it,” DiBona said. “But it
is kind of intangible.”
Google can measure the recruiting angle of how many students ended up
working for Google but that is only one possible metric that doesn’t
encompass the scope of what Google is trying to achieve.
“It’s very hard to say that this student made this project better that we
then used and made Google.com better which then led to advertising,” DiBona
admitted. “You can’t track that.”
He went on to explain that Google thinks that open source is important, and
that the program is a good way of supporting open source.
“We made the strategic decision to support open source, and this is one of
the ways we do it,” DiBona stated. “How you then try and determine ROI, I
don’t think that’s even possible.”
In 2005, only 30 percent of students stuck with their projects after the Summer of
Code ended.
“We’d probably add a little more documentation to the program to make it
more clear to participants that active engagement with their development
community is part of the program,” Hawthorn said.
The problem that Hawthorn heard from some of the mentors was that it was
sometimes difficult or intimidating for students to deal with the community
at large instead of just their own mentor.
is able to collaborate with the community at large they’ll have a better
chance of continuing with the project after the summer is over.
According to Hawthorn it’s also likely that there will be a 2007 Summer of
Code.
internally at every level so it’s likely that Google will be doing it again
next year.
For DiBona, it all boils down to giving back to the community from which
Google benefits so much.
“I really like open source people and open source developers a lot
personally, they’re idealists their hearts are in a good place,” DiBona
said.
fresh blood to open source, and I just think it’s so cool to be able to do
that.
“Google does derive so much value from open source software, that it’s not
like this is the least we can do, cause I’m sure we do a lot less. But it’s
pretty terrific to be able to do that here.”