Some things to remember:

1.  You don't need GSOC to contribute to a free software project, but
it's a great way to get started.  Almost all of our students have
stuck around afterwards.  some examples:

Holger Lubitz wrote a great deal of x86 assembly for GSOC, then came
back later to write his Masters' thesis on fast small SIMD integer
frequency transforms, which gave x264 an overall speed boost of up to
15%.  He got paid well for that work, too.

David Conrad wrote a great deal of ARM NEON assembly for GSOC and is
now our ARM maintainer.

Dylan Yudaken continued to work on tasks related to his GSOC project
and now maintains the code that he wrote (weighted prediction stuff).

2.  In my experience, people tend to assume that free software
developers, especially "good" ones, are far beyond their league.  They
see someone with hundreds of commits and thousands of lines of code to
their name and assume there is some magical divide between them and
the developer.  There isn't.

I'm a Junior.  I got started in my Freshman year: I hadn't used C for
nearly a decade.  I had never written or read a line of assembly code
for any processor.  I had no idea how video or image compression
worked.  I didn't even know what a Fourier transform was.

Since then, I've written 1/5 of x264, designed Facebook's encoding
backend, interned at a cable/IPTV broadcasting company, done contract
work for over a dozen companies, mentored for GSOC twice, and become
the x264 project leader.

Getting into open source is, IMO, the single best thing a CS major can
do for themselves, both in terms of hobbies, job prospects, and
friendships.  Do it as soon as possible and don't look back.


Finally, for general questions about GSOC, feel free to drop by #gsoc
on Freenode.  And of course, if you have questions specifically for
me, drop by #x264dev on Freenode.